# jurniti — full agent-readable index > jurniti is the secure managed agent platform. Each agent runs in its own real Firecracker microVM with hardware-enforced KVM isolation — not a shared Docker container. Your model API keys live only inside your VM (BYOK); we never see, proxy, or mark up your spend. Runs on AWS EC2 nested-virtualization hosts in us-west-2. From $25/mo with a 30-day money-back guarantee. This file is generated from the same data the website renders, so it never drifts. Curated short index: https://www.jurniti.com/llms.txt ## Harnesses The harnesses are the open-source agent runtimes you can run inside a jurniti microVM. Each gets its own subscription, microVM, and dedicated CPU + RAM. # Claude Code on jurniti > Claude Code hosting on a real Firecracker microVM — your Claude subscription or API key, always-on terminal, isolated from your laptop and every other tenant. Managed by jurniti; upstream project: [Anthropic](https://github.com/anthropics/claude-code). Anthropic's agentic coding CLI. Terminal-first, BYO Claude subscription. ## What you get - Fresh Ubuntu microVM with Claude Code pre-installed and on your PATH. - Live in-browser terminal — log in with your Anthropic subscription (claude /login) right here, or bring an ANTHROPIC_API_KEY. - Persistent home — your CLAUDE.md, skills, slash commands, and agents survive every restart. - Snapshot your setup into a forkable template — share your agent config, never your credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container — one tenant per kernel. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/claude-code --- # Claude Desktop on jurniti > Official Claude Desktop streamed to your browser — Chat, Cowork, and visual Code on an isolated Linux microVM, not your laptop. Managed by jurniti; upstream project: [Anthropic](https://code.claude.com/docs/en/desktop-linux). Anthropic's official desktop app (Chat, Cowork, Code), streamed from a Linux microVM. BYO Claude login. ## What you get - Official Anthropic Linux app (apt), already installed. Not a Windows repack. - KasmVNC desktop in the portal — close the tab, the guest stays up. - Sign in inside the desktop with your Claude subscription or org SSO. No Console API-key field. - Pro RAM floor (6 GiB). Electron + X11 does not fit Starter. - Mozilla Firefox as a .deb. Ubuntu's snap firefox is pinned out on purpose. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/claude-desktop --- # Codex CLI on jurniti > Run OpenAI Codex on a real microVM — the terminal-first coding agent, your ChatGPT subscription, fully isolated. Managed by jurniti; upstream project: [OpenAI](https://github.com/openai/codex). OpenAI's terminal-first coding agent. BYO ChatGPT subscription or API key. ## What you get - Fresh Ubuntu microVM with the Codex CLI pre-installed and on your PATH. - Live in-browser terminal — sign in with your ChatGPT account (codex login --device-auth for the headless flow) or bring an OPENAI_API_KEY. - Persistent home — your config, prompts, and AGENTS.md survive every restart. - Snapshot into a forkable template — share the setup, never the credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/codex --- # DeepAgents Code on jurniti > Run LangChain's DeepAgents Code on a real microVM — bring any model (GLM-5.2, DeepSeek, Kimi) via OpenRouter, no frontier-lab subscription required. Managed by jurniti; upstream project: [LangChain](https://github.com/langchain-ai/deepagents). LangChain's batteries-included, model-agnostic terminal coding agent. BYOK any provider. ## What you get - Fresh Ubuntu microVM with DeepAgents Code (dcode) pre-installed and on your PATH. - Truly model-agnostic — point dcode at any provider:model string (OpenRouter, Anthropic, OpenAI, Fireworks, Baseten, local). - Live in-browser terminal — dcode is a terminal-first TUI, no web UI to wrangle. - One-fork GLM-5.2 starter template: an open-weight coding agent for cents, not a subscription. - Snapshot into a forkable template — share the model wiring, never the credentials (scrubbed on publish). Real Firecracker isolation. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/deepagents --- # OpenCode on jurniti > Run OpenCode on a real microVM — the open-source, provider-agnostic terminal coding agent that works with any model. Managed by jurniti; upstream project: [SST](https://github.com/sst/opencode). Open-source, provider-agnostic terminal coding agent. Works with any model. ## What you get - Fresh Ubuntu microVM with OpenCode pre-installed and on your PATH. - Live in-browser terminal — configure any provider with opencode auth login. - Persistent home — your config and custom agents survive every restart. - Snapshot into a forkable template — share the setup, never the credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/opencode --- # Goose on jurniti > Run Block's Goose on a real microVM — the open, extensible on-machine AI agent that installs, edits, runs, and tests. Managed by jurniti; upstream project: [Block](https://github.com/block/goose). Block's open, extensible on-machine AI agent. Installs, edits, runs, tests. ## What you get - Fresh Ubuntu microVM with the Goose binary pre-installed and on your PATH. - Live in-browser terminal — set up any provider with goose configure. - Persistent home — your recipes and config survive every restart. - Snapshot into a forkable template — share the setup, never the credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/goose --- # Grok Build on jurniti > Run xAI's Grok Build on a real microVM — the terminal coding agent, your Grok subscription, isolated and disposable. Managed by jurniti; upstream project: [xAI](https://x.ai/cli). xAI's terminal coding agent. BYO Grok subscription or xAI API key. ## What you get - Fresh Ubuntu microVM with Grok Build pre-installed and on your PATH. - Live in-browser terminal — authenticate with your xAI subscription, or bring a GROK_CODE_XAI_API_KEY. - Persistent home — your agents and commands survive every restart. - Snapshot into a forkable template — share the setup, never the credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/grok --- # fx on jurniti > Run fx — Vercel Labs' 6MB Zig coding agent — on a real microVM. Unix-shell form factor, not an IDE in the terminal. Managed by jurniti; upstream project: [Vercel Labs](https://github.com/vercel-labs/fx). Vercel Labs' tiny Zig coding agent. Unix-shell form factor, BYO Vercel login or AI Gateway key. ## What you get - Fresh Ubuntu microVM with the fx binary pre-installed and on your PATH (~6.4 MiB, no Node runtime). - Live in-browser terminal — `fx login` is Vercel device-code (PTY-safe), or `fx setup` / AI_GATEWAY_API_KEY. - Persistent home — settings, skills, and AGENTS.md survive every restart. - Snapshot into a forkable template — share MCP wiring and skills, never auth.json or api-key (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/fx --- # Pool on jurniti > Run Poolside's pool on a real microVM — the agent harness Laguna models were trained in, terminal-first, BYOK. Managed by jurniti; upstream project: [Poolside](https://poolside.ai/). Poolside's terminal coding agent for Laguna models. BYO Poolside / OpenRouter key. ## What you get - Fresh Ubuntu microVM with the pool binary pre-installed and on your PATH. - Live in-browser terminal — pool login (Poolside / OpenRouter) or export POOLSIDE_API_KEY. - Persistent home — settings, AGENTS.md, and project state survive every restart. - Snapshot into a forkable template — share MCP wiring and permissions, never credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/pool --- # Prime Agent on jurniti > Run Prime Agent on a real microVM — Prime Intellect's self-improving RLM coding agent, always-on. Managed by jurniti; upstream project: [Prime Intellect](https://github.com/PrimeIntellect-ai/prime-agent). Self-improving RLM coding agent. IPython tools, continual harness, BYOK. ## What you get - Fresh Ubuntu microVM with Prime Agent pre-installed via the official installer. - Live in-browser terminal — /login or export your model key and start. - IPython-backed RLM tools, continual harness state, daemon-backed sessions. - Persistent ~/.prime/agent settings and skills; secrets scrubbed on public templates. - Real Firecracker isolation — not a shared container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/prime-agent --- # Pi on jurniti > Run Pi (the minimal coding-agent CLI) on a real microVM — 4-tool core, self-extending, JS-native. Managed by jurniti; upstream project: [Mario Zechner](https://github.com/badlogic/pi-mono). Minimal coding-agent CLI: 4-tool core (read/write/edit/bash) that self-extends. BYOK. ## What you get - Fresh Ubuntu microVM with Node + the upstream Pi installer ready to run. - Live in-browser terminal — Pi is terminal-first, no web UI to wrangle. - Persistent workspace — your skills and the agent's self-extensions survive restarts. - BYOK: bring your Anthropic / OpenAI / OpenRouter key. - Real Firecracker microVM isolation. Lightweight by default — even Starter (1 vCPU / 2 GiB) is plenty for Pi. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/pi --- # Gemini CLI on jurniti > Run Google's Gemini CLI on a real microVM — official coding agent, your Google auth or API key, isolated from your laptop. Managed by jurniti; upstream project: [Google](https://github.com/google-gemini/gemini-cli). Google's official Gemini CLI coding agent. Terminal-first, BYOK / Google auth. ## What you get - Fresh Ubuntu microVM with Gemini CLI pre-installed and on your PATH. - Live in-browser terminal — authenticate with Google or GEMINI_API_KEY right here. - Persistent home — settings and project context survive every restart. - Snapshot into a forkable template — share config, never credentials (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/gemini-cli --- # Cline on jurniti > Run the Cline agent runtime on a real microVM — open coding agent CLI, not a hosted VS Code extension. Managed by jurniti; upstream project: [Cline](https://github.com/cline/cline). Open coding-agent CLI (and VS Code extension). Host the agent runtime, not the IDE. ## What you get - Fresh Ubuntu microVM with the Cline CLI pre-installed. - Terminal-first runtime — connect your workflow; we host the agent box, not the IDE UI. - Live in-browser terminal — cline auth + interactive sessions. - Persistent home for rules and agent state; credentials scrubbed on public templates. - Real Firecracker isolation per tenant. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/cline --- # Kilo Code on jurniti > Run the Kilo Code agent on a real microVM — the open coding CLI across 500+ models, not an editor extension. Managed by jurniti; upstream project: [Kilo](https://github.com/kilo-org/kilocode). Open agentic coding CLI across 500+ models. Host the terminal agent, not the editor extension. ## What you get - Fresh Ubuntu microVM with the Kilo Code CLI (`kilo`) pre-installed. - Terminal-first runtime — `kilo auth login` in the browser terminal, BYOK across 500+ models. - Run one-shot tasks or full interactive sessions; swarm dispatch runs them in parallel. - Persistent home for agent rules and config; credentials scrubbed on public templates. - Real Firecracker isolation per tenant. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/kilocode --- # GitHub Copilot CLI on jurniti > Run GitHub Copilot CLI on a real microVM — your Copilot subscription, isolated terminal agent. Managed by jurniti; upstream project: [GitHub](https://docs.github.com/en/copilot). GitHub Copilot in the terminal. BYO GitHub / Copilot subscription. ## What you get - Fresh Ubuntu microVM with GitHub Copilot CLI on PATH. - Live in-browser terminal — log in with your GitHub / Copilot account. - Persistent home survives restarts and reprovisions. - Snapshot + fork your setup without leaking tokens (scrubbed on publish). - Real Firecracker microVM isolation. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/copilot --- # OpenHands on jurniti > Run OpenHands on a real microVM — the open software-development agent community expects after Claude Code / Codex. Managed by jurniti; upstream project: [OpenHands](https://github.com/OpenHands/OpenHands-CLI). OpenHands software development agent CLI. Terminal-first, BYOK LLM. ## What you get - Fresh Ubuntu microVM with the OpenHands CLI pre-installed. - Live in-browser terminal — configure your LLM and start coding. - Persistent ~/.openhands settings; sessions scrubbed on public templates. - Headless mode for automation; interactive TUI for day-to-day work. - Real Firecracker isolation — not a shared container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/openhands --- # Aider on jurniti > Run Aider on a real microVM — the classic terminal pair-programmer, BYOK any model. Managed by jurniti; upstream project: [Aider-AI](https://github.com/Aider-AI/aider). AI pair programming in your terminal. BYOK any model provider. ## What you get - Fresh Ubuntu microVM with Aider pre-installed via a pinned release. - Live in-browser terminal — export your API key and start pairing. - Git-native workflow: aider edits and commits with you. - Persistent conventions and config; secrets scrubbed on public templates. - Real Firecracker microVM isolation. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/aider --- # Amp on jurniti > Run Sourcegraph Amp on a real microVM — frontier coding agent, terminal-first, your Amp account. Managed by jurniti; upstream project: [Sourcegraph](https://ampcode.com). Sourcegraph Amp coding agent. Terminal-first, BYO Amp account. ## What you get - Fresh Ubuntu microVM with the Amp CLI installed. - Live in-browser terminal — sign in and start agents. - Persistent config home; credentials scrubbed on public templates. - BYO Amp subscription — we never resell model tokens. - Real Firecracker microVM isolation. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/amp --- # Cursor Agent on jurniti > Run Cursor Agent CLI on a real microVM — the same agent as Cursor, terminal-first, always-on, isolated from your laptop. Managed by jurniti; upstream project: [Anysphere](https://cursor.com/cli). Cursor Agent CLI — terminal coding agent. BYO Cursor account or API key. ## What you get - Fresh Ubuntu microVM with Cursor Agent (`agent`) pre-installed and on your PATH. - Live in-browser terminal — log in with NO_OPEN_BROWSER=1 agent login, or export CURSOR_API_KEY. - Persistent home — cli-config, rules, and project context survive every restart. - Snapshot into a forkable template — share the setup, never auth.json (scrubbed on publish). - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/cursor --- # Devin on jurniti > Run Devin for Terminal on a real microVM — Cognition's local coding agent with optional cloud handoff, always-on, isolated. Managed by jurniti; upstream project: [Cognition](https://cognition.com). Devin for Terminal — Cognition's local coding agent with cloud handoff. BYO Devin account. ## What you get - Fresh Ubuntu microVM with the Devin CLI pre-installed and on your PATH. - Live in-browser terminal — auth with `devin auth login --force-manual-token-flow` (PTY-safe, no localhost browser). - Persistent ~/.config/devin — skills, MCP wiring, and settings survive restarts. - Snapshot into a forkable template — credentials.toml scrubbed on public publish. - Real Firecracker microVM isolation. One tenant per kernel. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/devin --- # Hermes on jurniti > Run the Nous Research Hermes Agent on a real microVM — fresh Ubuntu, persistent home, your dashboard at .jurniti.com. Managed by jurniti; upstream project: [Nous Research](https://github.com/NousResearch/hermes-agent). Self-improving autonomous agent. Long-running, model-agnostic, BYOK. ## What you get - Fresh Ubuntu microVM with uv pre-installed and the upstream Hermes quickstart ready to copy-paste. - Public TLS subdomain for your Hermes Dashboard (auto-provisioned). - Persistent /home/hermes — your skills and memory survive every restart. - BYOK: paste your OpenRouter / Nous Portal / OpenAI key once. We never see your tokens. - Snapshot + restart, daily backups, real Firecracker isolation. Not a Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/hermes --- # Headlong on jurniti > Headlong keeps thinking between messages — a Laude/MIT Bash microharness on its own Firecracker guest. ~$1–2/hour while the mind is up. BYOK. Managed by jurniti; upstream project: [Laude Institute](https://github.com/laude-institute/headlong). Persistent-agency Bash microharness. The agent keeps thinking between messages. BYOK. ## What you get - Persistent agency: messages land in a thought stream that is already running. The agent decides if and when to reply. - Live in-browser terminal — run `headlong-init` with YOUR OpenRouter/Anthropic/OpenAI key. Set a provider spend cap first. - Measured spend while the mind is up: about $1–2/hour (~$720–1,440/month if left 24/7). `persona stop` is the off switch. - Persistent ~/.headlong — identity, memories, and trajectories survive restart. Public templates stay vetoed until a capture canary. - Real Firecracker isolation. Headlong's own sandbox is off on purpose: the microVM is the boundary. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/headlong --- # Exo on jurniti > Exo is a recursively self-improving agent harness — it can rewrite its own code, and those edits live on persist. BYOK. No vendor relay. Managed by jurniti; upstream project: [exoharness](https://github.com/exoharness/exo). Recursively self-improving agent harness. Rewrites its own code on persist. BYOK. ## What you get - Pinned checkout pre-baked. Write your OpenAI or OpenRouter key into ~/exo/byok.env, then run `exo`. - Self-improvements (SELF.md and the TypeScript harness the agent rewrites) live on persist — a reprovision keeps them. - Vendor ExoChat relay is off. The shipped config names no vendor host; a fork never inherits a relay identity. - Live in-browser terminal. No guest-bound dashboard to proxy. - Real Firecracker isolation, Docker sandbox in-guest. Public templates stay vetoed until a capture canary. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/exo --- # Hermes Desktop on jurniti > Official Hermes Agent Desktop streamed to your browser — the same agent core as the CLI, on an isolated Linux microVM, not your laptop. Managed by jurniti; upstream project: [Nous Research](https://hermes-agent.nousresearch.com/docs/user-guide/desktop). Official Hermes Agent desktop app, streamed from a Linux microVM. Same ~/.hermes as the CLI. BYOK. ## What you get - Official Nous Hermes Desktop (Electron), already built at bake. Not a web clone of the dashboard. - Same ~/.hermes as the CLI — keys, sessions, skills, and memory. - KasmVNC desktop in the portal — close the tab, the guest stays up. - BYOK OpenRouter or Nous Portal. Set a key once; both surfaces see it. - Pro RAM floor (6 GiB). Electron + X11 does not fit Starter. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/hermes-desktop --- # Unsloth Desktop on jurniti > Official Unsloth Desktop streamed to your browser — train and chat from an isolated Linux microVM, not your laptop. No GPU on this VM. Managed by jurniti; upstream project: [Unsloth](https://unsloth.ai/docs/desktop). Official Unsloth Desktop (Tauri) streamed from a Linux microVM. Cloud/remote or small CPU GGUFs — this VM has no GPU. ## What you get - Official Unsloth Ubuntu .deb (Tauri, unsloth-studio), already installed. Not the browser Studio on :8888. - KasmVNC desktop in the portal — close the tab, the guest stays up. - Use Cloud / Remote endpoints or a small CPU GGUF. Local 27B train needs a GPU this box does not have. - Sign in or set a Hugging Face token inside the desktop. No platform API-key field. - Pro RAM floor (6 GiB). Tauri + X11 does not fit Starter. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/unsloth-desktop --- # OpenClaw on jurniti > OpenClaw hosting on a real Firecracker microVM — Discord / WhatsApp / Signal / Telegram bridges, BYOK, always-on, your keys stay in the guest. Managed by jurniti; upstream project: [OpenClaw Foundation](https://github.com/openclaw/openclaw). Bridges Discord / WhatsApp / Signal / Telegram to your LLM. BYOK. ## What you get - Dedicated OpenClaw host: fresh Ubuntu microVM with the upstream installer ready to run. - Public TLS subdomain for the OpenClaw web UI (auto-provisioned). - Persistent workspace — conversations, plugin state, and skills survive restarts. - BYOK for every model provider OpenClaw supports (Claude, GPT, OpenRouter, local). - Real Firecracker isolation — no shared kernel with other tenants' OpenClaw bots. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/openclaw --- # Paperclip on jurniti > Run Paperclip on a real microVM — the open-source app for managing a team of AI agents at work, with its own web dashboard. Managed by jurniti; upstream project: [Paperclip](https://github.com/paperclipai/paperclip). Open-source app to manage a team of AI agents at work. Web UI, BYOK. ## What you get - Fresh Ubuntu microVM with Paperclip + an embedded database, pre-installed and ready. - Public TLS subdomain for the Paperclip dashboard (auto-provisioned) — manage your agent team from the browser. - Persistent workspace — your agents, workflows, and history survive every restart. - BYOK for every model provider, configured in the dashboard. We never see your keys. - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/paperclip --- # Mastra on jurniti > Run Mastra on a real microVM — the TypeScript agent framework, pre-built and served with its Studio web UI. Managed by jurniti; upstream project: [Mastra](https://github.com/mastra-ai/mastra). TypeScript agent framework — agents, tools & workflows with the Studio web UI. BYOK. ## What you get - Fresh Ubuntu microVM with a runnable Mastra project (agents, tools, workflows) pre-built and serving. - Public TLS subdomain for Mastra Studio (auto-provisioned) — chat with and test your agents from the browser. - Edit the TypeScript source in the built-in terminal, then one command rebuilds and redeploys: mastra-rebuild. - BYOK — one OpenRouter key unlocks hundreds of models via Mastra's model router. We never see your keys. - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/mastra --- # OpenScience on jurniti > Run OpenScience on a real microVM — the open-source AI workbench for scientific research, with its full browser workspace. Managed by jurniti; upstream project: [OpenScience](https://github.com/synthetic-sciences/openscience). Open-source AI workbench for scientific research — literature, code, experiments, write-up. BYOK. ## What you get - Fresh Ubuntu microVM with OpenScience pre-installed — research, ml, biology, and physics specialist agents ready. - Public TLS subdomain for the OpenScience workspace (auto-provisioned) — literature, code, experiments, and write-up in one browser UI. - Persistent home — projects, sessions, and skills survive every restart and can be snapshotted into templates. - BYOK for Anthropic, OpenAI, Gemini, and 75+ providers. Keys stay inside your microVM; we never proxy model spend. - Real Firecracker microVM isolation. Not a shared Docker container. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/openscience --- # QM on jurniti > Run YC's QM on a real microVM — multiplayer org agent with web UI, memory, crons, and sandboxes, without the Fly/AWS deploy tax. Managed by jurniti; upstream project: [YC Software](https://github.com/yc-software/qm). YC's multiplayer agent harness for work — org web UI, memory, crons, sandboxes. BYOK. ## What you get - Fresh Ubuntu microVM with the full QM stack (core + web-ui + Postgres) pre-baked and ready. - Public TLS subdomain for the QM web UI — portal auto-sign-in via the identity bridge (no Resend/SMTP wiring). - Persistent workspace + Postgres on /dev/vdb — org memory, files, and sessions survive every restart. - BYOK for the models QM drives (Pi / OpenCode / Codex / Claude Code). Keys stay inside your microVM. - Real Firecracker microVM isolation. Multi-container stack, one tenant per kernel. Pro plan floor. ## Pricing - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. 30-day money-back guarantee. No free trial, no free tier. Provisions in ~3 minutes after Stripe checkout. Product page: https://www.jurniti.com/qm --- # jurniti pricing > Flat fee per agent VM, billed monthly or annual (annual = 2 months free). BYOK for every model provider — no token markup. 30-day money-back guarantee, no free trial, no free tier. ## Plans - **Starter** — $25/mo or $250/yr (1 vCPU, 2 GiB RAM, 10 GiB SSD, 100 Mbit/s burst). Solo builders running Pi or Hermes 8B with a local proxy. Just enough to keep your agent's home alive 24/7. - **Pro** — $49/mo or $490/yr (2 vCPU, 6 GiB RAM, 50 GiB SSD, 200 Mbit/s burst, Custom subdomain). The default. Headroom for Hermes-3-8B at Q4, OpenClaw with a heavy plugin set, or a Pi instance you actually live in. - **Max** — $99/mo or $990/yr (4 vCPU, 12 GiB RAM, 100 GiB SSD, 500 Mbit/s burst, Daily snapshot). For agents you'd cry about losing. Daily snapshots, room for a 12 GiB working set, and the headroom to keep multiple skills warm. ## Templates plan (private templates) Public templates are unlimited and free forever. Private templates use a count-based plan — 4 tiers t5/t10/t20/t50 = 5/10/20/50 private slots at $5/$9/$17/$29 per month (additive 50-slot blocks past t50). There is no free private tier. Full pricing page: https://www.jurniti.com/pricing --- # jurniti docs > Developer docs for jurniti — the secure managed agent platform. Drive it from the terminal, CI, or your own agent with one CLI binary that doubles as an MCP server. Every command is also an MCP tool and a REST noun — one vocabulary. This index is agent-copyable; fetch the linked pages for detail. ## Docs - [Docs home](https://www.jurniti.com/docs): What is jurniti, building blocks (microVM / harness / template / plan), three lanes (quickstart, surfaces, reference). Dashboard + CLI + MCP. ([markdown](https://www.jurniti.com/docs.md)) - [Pricing](https://www.jurniti.com/pricing): Flat per-VM plans, BYOK for every model, 30-day money-back. No free trial or free tier. ([markdown](https://www.jurniti.com/pricing.md)) - [Templates gallery](https://www.jurniti.com/templates): Forkable snapshots of configured agent VMs. Fork one to provision a new VM seeded with it. ([markdown](https://www.jurniti.com/templates.md)) - [Full agent index](https://www.jurniti.com/llms-full.txt): The complete machine-readable site index — every harness, the pricing detail, and the public templates catalog. ## Docs pages - [Quickstart](https://www.jurniti.com/docs/quickstart): First agent in ~5 min — dashboard (magic link + Add agent) or CLI (login, add-card, up). Card entry is web-only via Stripe; no free tier, 30-day money-back on first sub. ([markdown](https://www.jurniti.com/docs/quickstart.md)) - [Dashboard](https://www.jurniti.com/docs/dashboard): Web path: magic-link login, Add agent wizard, browser terminal, harness portal, templates, Plan & billing (Stripe portal). No CLI required. ([markdown](https://www.jurniti.com/docs/dashboard.md)) - [Installation](https://www.jurniti.com/docs/installation): Install the CLI only if you want terminal/CI/MCP. One line: curl (macOS/Linux) or PowerShell (Windows). Prebuilt macOS, Linux, and Windows binaries. Dashboard-only users can skip. ([markdown](https://www.jurniti.com/docs/installation.md)) - [Usage tiers & credits](https://www.jurniti.com/docs/usage-tiers): Pay per-hour from prepaid credits — On-Demand and Spot for bursty agents. Top up on Plan → #credits; launch usage-tier VMs is CLI-first (jurniti up --tier). Spot is the swarm substrate. ([markdown](https://www.jurniti.com/docs/usage-tiers.md)) - [Token usage](https://www.jurniti.com/docs/token-usage): Advisory BYOK model-token meter (session/week, est. cost) on dashboard, jurniti token-usage, and MCP token_usage / vm_token_usage. Not fleet usage or compute credits; not a jurniti charge. ([markdown](https://www.jurniti.com/docs/token-usage.md)) - [Swarm](https://www.jurniti.com/docs/swarm): Fan out N agent microVMs on prepaid Spot credits, dispatch a task to each (jurniti run), and collect every result (jurniti cp). Setup -> dispatch -> collect; billed per second, one OpenRouter key across the swarm. CLI-first today. ([markdown](https://www.jurniti.com/docs/swarm.md)) - [Memory](https://www.jurniti.com/docs/memory): Attach hosted memory (Obsidian/gbrain/mem0) to agents from the dashboard, CLI (jurniti memory), or MCP (mcp_connection_grant) — buy in the UI, grant/revoke on every surface. ([markdown](https://www.jurniti.com/docs/memory.md)) - [Observability](https://www.jurniti.com/docs/observability): Attach Latitude/Langfuse sidecars to agents from the dashboard, CLI (jurniti observability), or MCP (observability_grant) — one active observability grant per agent in v1. ([markdown](https://www.jurniti.com/docs/observability.md)) - [Communication](https://www.jurniti.com/docs/communication): Buy a self-hosted Buzz relay in the dashboard, join as a human via Buzz desktop (Connect desktop), and grant agents from the relay card, jurniti communication, or MCP communication_grant (buzz-acp on wired harnesses). ([markdown](https://www.jurniti.com/docs/communication.md)) - [Agent skills & updates](https://www.jurniti.com/docs/agent-skills): Install jurniti's Agent Skills (open SKILL.md format, bundled in the binary) into Claude Code or Codex with jurniti skills install, and keep the CLI + skills current with the signed jurniti upgrade flow. Opt out of update notices with JURNITI_NO_UPDATE_CHECK=1. ([markdown](https://www.jurniti.com/docs/agent-skills.md)) - [CLI reference](https://www.jurniti.com/docs/cli-reference): Every verb, generated from the binary's own command tree. Money-touching: add-card, up, fork, credits buy/refund. ([markdown](https://www.jurniti.com/docs/cli-reference.md)) - [MCP](https://www.jurniti.com/docs/mcp): Mount the CLI as an MCP server; your agent gets the same verbs as tools and inherits your billing gate. ([markdown](https://www.jurniti.com/docs/mcp.md)) - [REST API](https://www.jurniti.com/docs/api): Customer REST at api.jurniti.com — OpenAPI at /openapi.json, bearer jrn_live_ keys, JSON {error, message} envelope. Same verbs as CLI and MCP. ([markdown](https://www.jurniti.com/docs/api.md)) ## CLI commands - `jurniti login` — authenticate this machine via the device flow (browser-optional) - `jurniti whoami` — show the identity the active API key resolves to - `jurniti harnesses` — list the available agent types and each one's auth requirement - `jurniti up --plan

--harness [--cycle monthly|annual] | --tier ondemand|spot --harness [--count N --tier spot] [--group NAME]` — provision a paid VM (flat plan charges your saved card; a usage tier meters hourly from prepaid credits). --count N fans out N agents at once (a swarm) — swarm agents are Spot-only, so --count requires --tier spot; --group NAME saves the swarm as a cohort for run/cp --group - `jurniti auth [--set K=V | --login]` — show how to make a VM work (set a BYOK key or one-time login). --set writes a BYOK key; --login opens a terminal on the VM and runs the agent's own interactive sign-in (e.g. a Claude subscription), so no credential is ever stored outside the VM - `jurniti shell [--login]` — open an interactive terminal inside your agent (Ctrl-] to leave). --login runs the agent's own sign-in command for you; the browser opens locally and the credential is minted inside the VM - `jurniti stop ` — pause a VM (persist preserved) - `jurniti start ` — resume a stopped VM - `jurniti restart ` — bounce a VM (reset rootfs, keep your data) - `jurniti reprovision ` — rebuild the guest, preserving your data - `jurniti capture --public|--private [--name ] [--memory ] [--observability ]` — snapshot a VM into a template (optional empty companion recipes) - `jurniti publish --public|--unlist` — flip a template's visibility - `jurniti fork [--plan

]` — provision a paid VM from a template (charges card on file; no free fork). Accepts template id or public/owned slug - `jurniti run "" | jurniti run --group [--var K=v1,v2] --model --provider

""` — dispatch one headless task to a VM's agent, or --group to a whole swarm at once (swarm Dispatch). With --group the task is a template: {placeholder} tokens fill per-agent from --var lists (KEY=v1,v2 or KEY=@file, cycled across the swarm) plus built-ins {index}/{n}/{vm}, so each agent does varied work. --model picks a per-run model (e.g. an OpenRouter slug) for the whole swarm without per-VM config (--provider names its provider, e.g. openrouter) — REQUIRED with --group because swarm slices have no default model; for a single VM, empty = its own configured default. --json (with --group) emits per-agent results for CI. Results land at work/out.txt — collect with cp - `jurniti cp : | jurniti cp --group ` — copy a file out of your VM, or --group to fan a whole swarm's output into a local dir (one file per agent) — swarm Collect. --json (with --group) emits per-agent results for CI - `jurniti wait | jurniti wait --group [--ready]` — block until a VM (or every agent in a swarm --group) reaches running, or --ready for harness_ready (dispatch-ready — you can `run` against it). Exits 0 when all are ready, non-zero on timeout. Fills the gap between up and run so you don't hand-roll a watch loop - `jurniti transfer [--vm ] [--harness ] [--map =] [--exclude ] [--dry-run] [--yes]` — move your local agent's sessions, memory, and config into your VM — losslessly, except credentials, which never leave this machine (you re-enter keys in the VM). Detects installed agents, shows a manifest of what's kept and what's scrubbed, rewrites path-keyed session stores so history resumes, and streams the archive straight into your VM (never stored on our servers). --dry-run previews the manifest and exits; with no VM yet, it walks you through checkout first - `jurniti vms ls` — list your VMs - `jurniti vms get ` — show one VM's detail - `jurniti templates ls [--q ] [--harness ] [--sort recent|stars|forks|trending] [--limit N] [--offset N]` — list your templates and the public gallery (optional search/filter). The gallery is paged: `total` is the whole matching catalog, so when the footer says there is more, walk it with --offset - `jurniti templates get ` — show one template's detail (skills, harness, visibility) - `jurniti usage` — show live VM counts and monthly spend (fleet/MRR — not model tokens; use token-usage for BYOK est.) - `jurniti token-usage [--window session|week] [--vm ]` — advisory model-token meter (BYOK est. spend by provider/model; not a jurniti charge). Distinct from `usage` (fleet/MRR). Optional --vm scopes to one owned box - `jurniti billing ` — show whether a card is on file, or add one (opens a browser) - `jurniti credits [buy |history|refund ]` — show your prepaid credit balance/burn/runway, buy credits ($10 minimum, charges your saved card), list the ledger, or refund unused balance - `jurniti keys ` — manage your tenant API keys (mint read-only or read-write keys) - `jurniti env | jurniti env set --group KEY=VALUE` — manage in-box env / BYOK config on one VM, or set the same key on every agent in a swarm group (--group). Values are written through to the guest and never echoed back - `jurniti automations ` — manage curated loop packs on your agents: list packs, enable/pause/remove, run now, history, and review queue - `jurniti memory ` — list memory/MCP connections and grant or revoke them onto agent VMs (same outcomes as the dashboard Memories page) - `jurniti observability ` — list observability sidecar grants and attach or detach a Latitude/Langfuse sidecar onto an agent VM - `jurniti communication ` — list Buzz communication relays and grant or revoke agent attach (buzz-acp inject; buy relay from the dashboard) - `jurniti skills ls` — list the bundled agent skill playbooks - `jurniti skills install [--target claude|codex] [--dir ]` — install the bundled skills into your agent (Claude Code or Codex). Skills ship embedded in the binary (no network); installed copies are version-stamped and recorded so `jurniti upgrade` / `skills sync` keep them fresh - `jurniti skills sync` — bring previously installed skills up to this CLI's bundled content (runs automatically after `jurniti upgrade`). Only overwrites files carrying the jurniti version stamp — user-modified skills are left alone; never rolls a newer stamp backward - `jurniti upgrade [--check] [--json] [--yes]` — self-update the CLI to the latest signed release (checksum + minisign verified, atomic swap, then re-syncs installed skills). --check only reports status — exit 0 whether up to date, update available, or the check failed (the signal is in the JSON fields, so agents and CI can run it unconditionally). Executing requires an interactive terminal unless --yes is passed. Passive update notices obey JURNITI_NO_UPDATE_CHECK=1 and are suppressed in CI, non-TTY, and --json runs - `jurniti docs [--format=md|llms]` — print this CLI's command reference (md) or agent llms.txt index - `jurniti mcp` — run an MCP stdio server so your own agent drives your fleet ## Install - macOS / Linux: `curl -fsSL https://jurniti.com/install.sh | sh` - Windows (PowerShell): `irm https://jurniti.com/install.ps1 | iex` ## MCP - Mount jurniti as an MCP server for your agent: `claude mcp add jurniti -- jurniti mcp` ## Notes for agents - Card entry is web-only: `jurniti billing add-card` opens the browser to the jurniti card page, where details are typed into Stripe's embedded fields (Stripe Elements). The card goes straight to Stripe — the CLI and jurniti's servers never see or handle the number. - Your agent inherits your gate: when your agent calls `up`/`fork` over MCP it starts real, billed VMs under your card — the no-free-tier gate applies to an agent exactly as to a human. - Scope agent keys: mint a read-only key (`jurniti keys issue`) for an agent that should only observe, a read-write key for one that provisions; rotate or revoke with `jurniti keys rotate` / `jurniti keys revoke`. ## Skills > Copyable agent playbooks (SKILL.md) that teach your own Claude Code / Codex the jurniti workflows. Fetch one and drop it in your agent's skills dir, or run `jurniti skills install`. - [jurniti](https://www.jurniti.com/skills/jurniti.md): Orientation for the jurniti developer CLI and MCP surface — how to install it, log in, mount it into your own agent, and the full verb map. Use when the user mentions jurniti, wants to manage agent microVMs from the terminal or CI, asks how to use the jurniti CLI, or wants to give their own agent jurniti's tools. - [jurniti-fork-template](https://www.jurniti.com/skills/jurniti-fork-template.md): Fork a public or private jurniti template into a fresh paid VM and customize it — discover with templates ls, inspect with templates get, fork, wait until ready, then make it yours. Use when the user wants to fork a template, "clone a jurniti template", start from an existing agent setup, or bootstrap a VM from the gallery. - [jurniti-provision-agent](https://www.jurniti.com/skills/jurniti-provision-agent.md): Provision a fresh managed agent microVM on jurniti and connect to it — the log in, add a card, bring it up, connect flow. Use when the user wants to spin up a new agent VM, "provision an agent", "start a new jurniti box", or bring a fresh agent runtime online. - [jurniti-templatize](https://www.jurniti.com/skills/jurniti-templatize.md): Turn a configured jurniti VM into a reusable template — capture it public or private, then publish or flip its visibility. Use when the user wants to templatize a VM, "save this box as a template", "make a forkable snapshot", or share their agent setup. ## Getting a VM - Paid provision, no free tier: `jurniti login` → `jurniti billing add-card` (card on file) → `jurniti up`. Every plan ships a 30-day money-back guarantee instead of a free trial. --- # jurniti blog — guides > Long-form guides on running agent harnesses on jurniti microVMs — one per supported agent, comparisons included. Every post has an agent-copyable markdown twin — append `.md` to its URL. - [Headlong on a MicroVM: The Always-Thinking Agent](https://www.jurniti.com/blog/headlong-on-a-microvm) (2026-08-26) — Headlong is a Laude/MIT Bash microharness with persistent agency. Leave the mind running on a Firecracker guest. BYOK. About $1–2/hour while it thinks. ([markdown](https://www.jurniti.com/blog/headlong-on-a-microvm.md)) - [Exo on a MicroVM: The Harness That Rewrites Itself](https://www.jurniti.com/blog/exo-on-a-microvm) (2026-08-26) — Exo is a recursively self-improving agent harness. Run it always-on in a Firecracker guest. BYOK. Self-edits live on persist, not on an ephemeral rootfs. ([markdown](https://www.jurniti.com/blog/exo-on-a-microvm.md)) - [Remote CLI AI Coding Agent on a jurniti MicroVM](https://www.jurniti.com/blog/remote-cli-dev-planner-template) (2026-08-20) — Plan on this box. Get a worktree diff. Ask it to merge without the branch name. It refuses. Fork the Claude Code planner. Your key. ([markdown](https://www.jurniti.com/blog/remote-cli-dev-planner-template.md)) - [AI Account Expert Agent on a jurniti MicroVM](https://www.jurniti.com/blog/account-expert-template) (2026-08-20) — Ask about Acme. Get SSO and the 12th. Ask about Northwind. Get INV-1842. Files never mix. Fork the Hermes template. Your key. ([markdown](https://www.jurniti.com/blog/account-expert-template.md)) - [fx on a MicroVM: The 6MB Unix-like Coding Agent](https://www.jurniti.com/blog/fx-on-a-microvm) (2026-08-19) — fx is a 6.4 MiB Zig coding agent that behaves like a shell, not an IDE. Run it always-on in a Firecracker guest. BYOK, from $25/mo. ([markdown](https://www.jurniti.com/blog/fx-on-a-microvm.md)) - [AI Meeting Follow-Up Agent on a jurniti MicroVM](https://www.jurniti.com/blog/meeting-prep-follow-up-template) (2026-08-18) — Draft the follow-up from yesterday. Priya, the 12th, Friday security page. Nothing sends. Fork the Hermes meeting template. Your key. ([markdown](https://www.jurniti.com/blog/meeting-prep-follow-up-template.md)) - [AI Inbox Triage Agent on a jurniti MicroVM](https://www.jurniti.com/blog/inbox-triage-reply-digest-template) (2026-08-18) — Ask what needs me. Get Priya and Marcus — drafts in your voice, newsletter out, nothing sent. Fork the Hermes inbox template. Your key. 30-day money-back. ([markdown](https://www.jurniti.com/blog/inbox-triage-reply-digest-template.md)) - [AI Chief of Staff Agent on a jurniti MicroVM](https://www.jurniti.com/blog/chief-of-staff-fleet-router-template) (2026-08-18) — Ask what's in flight. Get Acme and INV-1842. Ask it to spin up a VM — it refuses. Fork the Hermes chief-of-staff template. Your key. 30-day money-back. ([markdown](https://www.jurniti.com/blog/chief-of-staff-fleet-router-template.md)) - [Unsloth Desktop on Linux: Isolated Always-On Box](https://www.jurniti.com/blog/unsloth-desktop-linux) (2026-08-17) — Install Unsloth Desktop on Linux (Ubuntu). Official Tauri app, always-on isolated box for studio workflows. CPU-only VM — not a GPU training box. ([markdown](https://www.jurniti.com/blog/unsloth-desktop-linux.md)) - [Hermes Agent Desktop: Isolated Always-On Box](https://www.jurniti.com/blog/hermes-agent-desktop) (2026-08-17) — Hermes Agent Desktop is the official GUI on the same ~/.hermes as the CLI. Isolated always-on box so the app stays up after you close the laptop. ([markdown](https://www.jurniti.com/blog/hermes-agent-desktop.md)) - [Claude Desktop vs Claude Code: Which Surface](https://www.jurniti.com/blog/claude-desktop-vs-claude-code) (2026-08-17) — Claude Desktop is the visual surface. Claude Code is the terminal. Here's when to use each, and when you want both on an isolated box. ([markdown](https://www.jurniti.com/blog/claude-desktop-vs-claude-code.md)) - [Claude Desktop Linux: Isolated Always-On Box](https://www.jurniti.com/blog/claude-desktop-linux) (2026-08-17) — Official Claude Desktop on Linux is apt plus a local session. Here's the isolated always-on path that keeps the GUI without tying it to your laptop. ([markdown](https://www.jurniti.com/blog/claude-desktop-linux.md)) - [Claude Code Skills That Earn Their Place](https://www.jurniti.com/blog/claude-code-skills) (2026-08-17) — We run 78 Claude Code skills in production, continuously. Which ones changed an outcome, which are shelfware, and how to tell before you install 100 of them. ([markdown](https://www.jurniti.com/blog/claude-code-skills.md)) - [AI Employees, Ranked by Evidence (Not Demos)](https://www.jurniti.com/blog/ai-employees) (2026-08-15) — AI employees ranked by published evidence: Kavak's 2.1x sales agents, an AI CEO up 50% profit, coding agents shipping 95% of PRs — and what they run on. ([markdown](https://www.jurniti.com/blog/ai-employees.md)) - [Claude Sync: Keep Claude Code Sessions on Every Device](https://www.jurniti.com/blog/sync-claude-code-sessions) (2026-08-13) — Claude Code doesn't sync sessions between machines. Where your history actually lives, why copying ~/.claude breaks resume, and how to move it losslessly. ([markdown](https://www.jurniti.com/blog/sync-claude-code-sessions.md)) - [How to Back Up OpenClaw by Hand (and Restore It Anywhere)](https://www.jurniti.com/blog/how-to-back-up-openclaw) (2026-08-13) — What an OpenClaw backup must include, a tar command that captures it all safely, the three ways restores break, and the one-command cloud restore. ([markdown](https://www.jurniti.com/blog/how-to-back-up-openclaw.md)) - [Prime Agent on a MicroVM: RLM Coding Without Shared Kernels](https://www.jurniti.com/blog/primeagent-on-a-microvm) (2026-08-06) — Prime Agent wants a real machine. jurniti runs it on a Firecracker tenant microVM — BYOK, always-on, from $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/primeagent-on-a-microvm.md)) - [Self-Host Buzz by Block — or Skip the VPS (Managed Install)](https://www.jurniti.com/blog/how-to-install-buzz) (2026-08-06) — Self-host Buzz by Block with Docker Compose, or skip the VPS: jurniti runs the Buzz relay in a Firecracker microVM with Open Buzz and agent grants. ([markdown](https://www.jurniti.com/blog/how-to-install-buzz.md)) - [Start an AI Agency Without Becoming Free Tech Support](https://www.jurniti.com/blog/start-an-ai-agency) (2026-08-03) — Start an AI agency with productized scope and a delivery stack. Free Agency 101 guide. No free tier; 30-day money-back. ([markdown](https://www.jurniti.com/blog/start-an-ai-agency.md)) - [Reddit Launch Playbook for Claude Code: One Sub a Day](https://www.jurniti.com/blog/reddit-launch-playbook-for-claude-code-template) (2026-08-03) — Map niche subs, one story, one post per day with UTM CTAs, 15-day refresh. Paid or money-back honesty — never free trial. Sibling to pure-value comment karma. ([markdown](https://www.jurniti.com/blog/reddit-launch-playbook-for-claude-code-template.md)) - [QM on a MicroVM: YC's Multiplayer Work Agent](https://www.jurniti.com/blog/qm-on-a-microvm) (2026-08-03) — YC's multiplayer work agent (QM) on a Firecracker microVM — org UI, memory, sandboxes. Pro plan, BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/qm-on-a-microvm.md)) - [Multiple AI Agents: Roles, Not One Oracle](https://www.jurniti.com/blog/multiple-ai-agents) (2026-08-03) — Multiple AI agents means role-scoped fleets with ownership—not one chat that does everything. OS design for AI-native companies. ([markdown](https://www.jurniti.com/blog/multiple-ai-agents.md)) - [Multi Agent Orchestration Platform Checklist](https://www.jurniti.com/blog/multi-agent-orchestration-platform) (2026-08-03) — A multi agent orchestration platform coordinates roles, tools, isolation, and cost—not just a DAG UI. Buyer checklist for AI-native fleets. ([markdown](https://www.jurniti.com/blog/multi-agent-orchestration-platform.md)) - [Managed AI Agents: Third Way Hosting](https://www.jurniti.com/blog/managed-ai-agents) (2026-08-03) — Managed AI agents mean always-on hosts without surrendering keys—microVMs, BYOK, templates. How AI-native companies host fleets. ([markdown](https://www.jurniti.com/blog/managed-ai-agents.md)) - [LLM Observability for Agent Fleets](https://www.jurniti.com/blog/llm-observability) (2026-08-03) — LLM observability means you can see what ran, what it cost, whether it finished, and why it failed—plus model token usage for BYOK fleets. ([markdown](https://www.jurniti.com/blog/llm-observability.md)) - [How to Start an AI Automation Agency (Week-One Checklist)](https://www.jurniti.com/blog/how-to-start-an-ai-automation-agency) (2026-08-03) — How to start an AI automation agency: one offer, baseline metrics, multi-agent map, first channel. Free 7-day Agency 101. ([markdown](https://www.jurniti.com/blog/how-to-start-an-ai-automation-agency.md)) - [How to Run an AI Agent Agency (Without Becoming the Pager)](https://www.jurniti.com/blog/how-to-run-an-ai-agent-agency) (2026-08-03) — Run an AI automation agency: client-math pricing, multi-agent fleets, memory, observability, channels, plugins. Free 7-day Agency 101 guide. ([markdown](https://www.jurniti.com/blog/how-to-run-an-ai-agent-agency.md)) - [How to Build an AI-Native Company](https://www.jurniti.com/blog/how-to-build-an-ai-native-company) (2026-08-03) — An AI-native company is an operating system—multi-agent fleets, observability, memory, communication channels, and plugins—not a chat widget. Free 7-day course. ([markdown](https://www.jurniti.com/blog/how-to-build-an-ai-native-company.md)) - [3D Educational AI Tutor on Codex: Curricula You Can Orbit](https://www.jurniti.com/blog/educational-3d-curriculum-tutor-template) (2026-08-03) — Fork a Codex microVM that builds multi-lesson 3D curricula — diagnose, curriculum.md, hotspots, quiz. Seed anatomy included. BYOK. No free trial. ([markdown](https://www.jurniti.com/blog/educational-3d-curriculum-tutor-template.md)) - [AI Services Agency: Productized Seats, Not Staff Aug](https://www.jurniti.com/blog/ai-services-agency) (2026-08-03) — Build an AI services agency as seats with outcomes—not staff-aug hours. Free Agency 101. Managed microVMs from $25/mo. ([markdown](https://www.jurniti.com/blog/ai-services-agency.md)) - [AI Consultant vs AI Automation Agency: Which Path Fits](https://www.jurniti.com/blog/ai-consultant-vs-ai-automation-agency) (2026-08-03) — AI consultant vs AI automation agency: role, pricing, delivery stack. Free Agency 101 guide. No free trial; 30-day money-back on first purchase. ([markdown](https://www.jurniti.com/blog/ai-consultant-vs-ai-automation-agency.md)) - [AI Automation Agency Pricing: Client Math, Not Hours](https://www.jurniti.com/blog/ai-automation-agency-pricing) (2026-08-03) — Price AI automation agency work from annualized client cost. Quit hourly after proof. Free Agency 101. No free trial. ([markdown](https://www.jurniti.com/blog/ai-automation-agency-pricing.md)) - [AI Agent Services: What to Sell (and What Not To)](https://www.jurniti.com/blog/ai-agent-services-what-to-sell) (2026-08-03) — Package AI agent services clients buy: audits, operators, retainers. Avoid mega-prompts. Free 7-day Agency 101. No free trial. ([markdown](https://www.jurniti.com/blog/ai-agent-services-what-to-sell.md)) - [AI Agent Security: Isolation and Blast Radius](https://www.jurniti.com/blog/ai-agent-security) (2026-08-03) — AI agent security is blast radius, BYOK key custody, and isolation—not a slide that says secure. How AI-native companies keep fleets safe. ([markdown](https://www.jurniti.com/blog/ai-agent-security.md)) - [AI Agent Observability for Agencies](https://www.jurniti.com/blog/ai-agent-observability-for-agencies) (2026-08-03) — AI agent observability turns retainers into evidence: runs, fails, traces. Free Agency 101. Link to self-host Langfuse. ([markdown](https://www.jurniti.com/blog/ai-agent-observability-for-agencies.md)) - [AI Agent Monitoring: Cost, Health, Honesty](https://www.jurniti.com/blog/ai-agent-monitoring) (2026-08-03) — AI agent monitoring covers health, cost (tokens), and failure modes—so multi-agent companies stay accountable. ([markdown](https://www.jurniti.com/blog/ai-agent-monitoring.md)) - [AI Agent Consulting: Delivery That Survives the Call](https://www.jurniti.com/blog/ai-agent-consulting) (2026-08-03) — AI agent consulting needs isolation, memory, and proof—not slide decks. Free Agency 101 for operators. 30-day money-back. ([markdown](https://www.jurniti.com/blog/ai-agent-consulting.md)) - [AI Agency Business Model: Build, Retainer, Expansion](https://www.jurniti.com/blog/ai-agency-business-model) (2026-08-03) — AI agency business model: fixed builds, maintenance retainers, expansion from baseline metrics. Free Agency 101 email guide. ([markdown](https://www.jurniti.com/blog/ai-agency-business-model.md)) - [Agentic AI Agency: Multi-Agent Delivery Without the Hype](https://www.jurniti.com/blog/agentic-ai-agency) (2026-08-03) — What an agentic AI agency actually ships: multi-agent roles, isolation, memory, observability. Free Agency 101. 30-day money-back. ([markdown](https://www.jurniti.com/blog/agentic-ai-agency.md)) - [Agent Tracing: Production Run Accountability](https://www.jurniti.com/blog/agent-tracing) (2026-08-03) — Agent tracing means structured run metadata—what tools ran, what failed, what it cost—so fleets are operable. ([markdown](https://www.jurniti.com/blog/agent-tracing.md)) - [Agent Swarm: Many Agents Without Company Collapse](https://www.jurniti.com/blog/agent-swarm) (2026-08-03) — An agent swarm is a fleet of scoped agents with blast radius—not one hero chat. How AI-native companies orchestrate swarms with isolation and cost honesty. ([markdown](https://www.jurniti.com/blog/agent-swarm.md)) - [Lead Magnet Ideas Are Cheap. This Agent Drafts the Funnel](https://www.jurniti.com/blog/lead-magnet-funnel-writer-template) (2026-08-02) — Fork a Claude Code microVM that drafts your lead magnet word-for-word, both funnel pages as HTML, break-even math, and six follow-up emails. It never sends. ([markdown](https://www.jurniti.com/blog/lead-magnet-funnel-writer-template.md)) - [5-Day Challenge Operator on Claude Code](https://www.jurniti.com/blog/5-day-challenge-operator-for-claude-code-template) (2026-08-02) — Fork a Claude Code microVM that designs your 5-day live challenge: curriculum, free+VIP funnel copy, emails, ops, and cart language. Draft-only. You go live. ([markdown](https://www.jurniti.com/blog/5-day-challenge-operator-for-claude-code-template.md)) - [Paid Ads → De-anon → Instantly + Plunk Drip (Hermes)](https://www.jurniti.com/blog/paid-ads-deanon-instantly-plunk-drip-for-hermes-template) (2026-07-30) — Non-converting paid traffic to Instantly cold drafts and open-source Plunk nurture drips. Warmly default, RB2B optional, human gates. Fork on Hermes. ([markdown](https://www.jurniti.com/blog/paid-ads-deanon-instantly-plunk-drip-for-hermes-template.md)) - [Reddit Karma Loop on a Forkable Claude Code MicroVM](https://www.jurniti.com/blog/loopany-reddit-karma-for-claude-code-template) (2026-07-30) — Loopany Episode 1 Reddit karma pattern for Claude Code: personal wiki, hard filters, account ledger, OpenCLI. Draft-first. No free trial. ([markdown](https://www.jurniti.com/blog/loopany-reddit-karma-for-claude-code-template.md)) - [Run Devin CLI on a Server (Own microVM)](https://www.jurniti.com/blog/devin-hosting) (2026-07-30) — Run Cognition's Devin CLI on its own Firecracker microVM — terminal-first, PTY-safe login, persistent home. From $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/devin-hosting.md)) - [Run Cursor Agent CLI as a 24/7 Server](https://www.jurniti.com/blog/cursor-hosting) (2026-07-30) — Run Cursor's agent CLI as an always-on server on a Firecracker microVM. BYO Cursor account or API key. Headless login, persistent home. From $25/mo. ([markdown](https://www.jurniti.com/blog/cursor-hosting.md)) - [Buzz ACP Hosting: Multi-Harness Agents](https://www.jurniti.com/blog/buzz-acp-multi-harness-agents) (2026-07-30) — Run Buzz ACP multi-harness agents on Firecracker microVMs — one shared room, isolated guests, always-on. From $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/buzz-acp-multi-harness-agents.md)) - [AI Video Ads in the Vox Paper-Collage Style: The Setup](https://www.jurniti.com/blog/vox-style-paper-collage-ad-template) (2026-07-29) — AI video ads in the Vox paper-collage style. Chat 30 seconds, approve once, get chained Omni Flash clips or paste-ready prompts. BYOK. Fork on jurniti… ([markdown](https://www.jurniti.com/blog/vox-style-paper-collage-ad-template.md)) - [Marketing AI Agent: How the Viral X Launch Operator Works](https://www.jurniti.com/blog/viral-x-launch-operator-marketing-skills-template) (2026-07-29) — A marketing AI agent that runs your X launch day: writes the kit, posts through Postiz, drafts your warm pings, works the first hour. You approve, it posts.… ([markdown](https://www.jurniti.com/blog/viral-x-launch-operator-marketing-skills-template.md)) - [AI Tutor on Codex: Adaptive Learning, Session by Session](https://www.jurniti.com/blog/teacher-adaptive-learn-tutor-codex-template) (2026-07-29) — AI tutor on Codex that teaches any topic adaptively, then writes an interactive HTML lesson you keep forever. Worked example, quiz, one file. BYOK. Fork on… ([markdown](https://www.jurniti.com/blog/teacher-adaptive-learn-tutor-codex-template.md)) - [Smartlead AI: Run a Cold Email Operator on a MicroVM](https://www.jurniti.com/blog/smartlead-cold-email-operator-template) (2026-07-29) — Smartlead AI on Claude Code: drop in your list, get a draft campaign, an opener per lead, and a warm-inbox check. Nothing auto-sends. You hit Start. Fork on… ([markdown](https://www.jurniti.com/blog/smartlead-cold-email-operator-template.md)) - [Sentry AI Skills for Claude Code: What the Pack Does](https://www.jurniti.com/blog/sentry-for-ai-skills-pack-sf-template) (2026-07-29) — Sentry AI skills on your own isolated Claude Code VM. First-party getsentry skills, your token, you approve every monitor change. 30-day money-back. Fork on… ([markdown](https://www.jurniti.com/blog/sentry-for-ai-skills-pack-sf-template.md)) - [Sentry AI Workflow: From Error to Draft PR](https://www.jurniti.com/blog/sentry-error-to-draft-pr-sf-template) (2026-07-29) — Sentry AI agent that turns a production error into a draft fix PR: stack summary, suspect commits, minimal patch. You review and merge. Nothing auto-ships.… ([markdown](https://www.jurniti.com/blog/sentry-error-to-draft-pr-sf-template.md)) - [Obsidian-Style Memory for AI Agents on jurniti](https://www.jurniti.com/blog/self-host-obsidian-memory-for-ai-agents) (2026-07-29) — Obsidian-Style Memory for AI Agents on jurniti. Isolated jurniti engine path for agents. BYOK. Paid plans. 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/self-host-obsidian-memory-for-ai-agents.md)) - [Latitude Observability for AI Agents on jurniti](https://www.jurniti.com/blog/self-host-latitude-for-ai-agents) (2026-07-29) — Latitude Observability for AI Agents on jurniti. Isolated jurniti engine path for agents. BYOK. Paid plans. 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/self-host-latitude-for-ai-agents.md)) - [Self-Host gbrain Memory for AI Agents](https://www.jurniti.com/blog/self-host-gbrain-for-ai-agents) (2026-07-29) — Self-Host gbrain Memory for AI Agents. Isolated jurniti engine path for agents. BYOK. Paid plans. 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/self-host-gbrain-for-ai-agents.md)) - [Buzz Communication Engine for AI Agents on jurniti](https://www.jurniti.com/blog/self-host-buzz-for-ai-agents) (2026-07-29) — Buzz Communication Engine for AI Agents on jurniti. Isolated jurniti engine path for agents. BYOK. Paid plans. 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/self-host-buzz-for-ai-agents.md)) - [SaaS BOFU SEO Content Agent on jurniti](https://www.jurniti.com/blog/saas-bofu-seo-content-engine-for-claude-code-template) (2026-07-29) — An AI SEO agent that drafts bottom-funnel comparison pages with CTAs built in. Checks page one first. Never auto-publishes. Fork it, boot in about 3 minutes.… ([markdown](https://www.jurniti.com/blog/saas-bofu-seo-content-engine-for-claude-code-template.md)) - [QuickBooks AI: Cash Flow + Invoice Chasing with an Agent](https://www.jurniti.com/blog/quickbooks-cash-flow-invoice-operator-template) (2026-07-29) — QuickBooks AI for 'will I make payroll?' - 30/60/90-day cash forecast from your real books + drafted overdue-invoice reminders. It drafts; you send. Fork on… ([markdown](https://www.jurniti.com/blog/quickbooks-cash-flow-invoice-operator-template.md)) - [Higgsfield AI Product Photography: Setup Without a Shoot](https://www.jurniti.com/blog/product-shoot-pack-higgsfield-template) (2026-07-29) — Higgsfield AI product photography without a shoot: studio, lifestyle, and marketplace images from photos you already have, filed for client delivery. Fork on… ([markdown](https://www.jurniti.com/blog/product-shoot-pack-higgsfield-template.md)) - [AI Social Media Scheduler: Postiz + OpenClaw Setup](https://www.jurniti.com/blog/postiz-social-media-scheduler-template) (2026-07-29) — Postiz social media scheduler run by OpenClaw: say what to post and when, and it drafts, queues, and publishes to 28+ channels from one calendar. BYOK. Fork… ([markdown](https://www.jurniti.com/blog/postiz-social-media-scheduler-template.md)) - [Paid Ads ROAS Board for Hermes — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/paid-ads-roas-board-for-hermes-template) (2026-07-29) — Ads dashboard with a budget gate: drafts hooks and CTAs to an Obsidian board you own, sets break-even ROAS before a dollar of spend. One channel, mastered.… ([markdown](https://www.jurniti.com/blog/paid-ads-roas-board-for-hermes-template.md)) - [Paid Ads → De-anon → Instantly — Hermes Cheat-Code Template](https://www.jurniti.com/blog/paid-ads-deanon-instantly-for-hermes-template) (2026-07-29) — Turn non-converting paid traffic into Instantly sequences and booked demos. Warmly default de-anon, RB2B optional, MillionVerifier, human gates. Fork on Hermes. ([markdown](https://www.jurniti.com/blog/paid-ads-deanon-instantly-for-hermes-template.md)) - [AI Ad Generator Workflow: Research, Angles, Batch, Log](https://www.jurniti.com/blog/paid-acquisition-agent-jockey-for-claude-code-template) (2026-07-29) — AI ad generator that reads its own results: research pains, stack angles, batch creatives, log winners. Nothing spends without you. From $25/mo. Fork on… ([markdown](https://www.jurniti.com/blog/paid-acquisition-agent-jockey-for-claude-code-template.md)) - [OpenAI Codex Security on a Forkable Codex MicroVM](https://www.jurniti.com/blog/openai-codex-security-for-codex-template) (2026-07-29) — OpenAI Codex Security CLI + skill pack, already on a managed Codex box. Dry-run first. Your keys. Your findings. No free trial. ([markdown](https://www.jurniti.com/blog/openai-codex-security-for-codex-template.md)) - [Nightly Antipattern Fixer Agent on jurniti](https://www.jurniti.com/blog/nightly-one-antipattern-fixer-sf-template) (2026-07-29) — Overnight code cleanup agent: finds one safe lint or type fix per night and opens one small draft PR. Never touches auth, billing, or migrations. Fork on… ([markdown](https://www.jurniti.com/blog/nightly-one-antipattern-fixer-sf-template.md)) - [Net Revenue Retention: The One Metric Your AI Can't Source](https://www.jurniti.com/blog/net-revenue-retention-stripe-template) (2026-07-29) — How to calculate net revenue retention from Stripe, why no single endpoint returns it, and how to get a weekly churn read where every figure is sourced. ([markdown](https://www.jurniti.com/blog/net-revenue-retention-stripe-template.md)) - [Apollo Alternative on Pay-Per-Lead: The Monid Setup](https://www.jurniti.com/blog/monid-apollo-pay-per-lead-template) (2026-07-29) — Apollo alternative inside Claude Code: pay about $0.05 per enriched B2B lead with verified work email. No $4,000/yr Apollo seat. Never auto-sends. Fork on… ([markdown](https://www.jurniti.com/blog/monid-apollo-pay-per-lead-template.md)) - [AI Monday Brief Operator on a jurniti MicroVM](https://www.jurniti.com/blog/monday-brief-operator-template) (2026-07-29) — QuickBooks + HubSpot in one Monday brief: cash on hand, stuck deals, and your top 3 moves before your coffee's cold. Read-only - it reports, you act. Fork on… ([markdown](https://www.jurniti.com/blog/monday-brief-operator-template.md)) - [Reddit Ad Research Agent: Mine the Angle on jurniti](https://www.jurniti.com/blog/mine-reddit-for-the-ad-angle-for-claude-code-template) (2026-07-29) — Reddit marketing tool for Claude Code: mine real complaints, rank the top 3 pains, write ads in your customers' words. Nothing spends without you. Fork on… ([markdown](https://www.jurniti.com/blog/mine-reddit-for-the-ad-angle-for-claude-code-template.md)) - [Higgsfield AI Ads: The Weekly Meta Creative Loop](https://www.jurniti.com/blog/meta-creative-loop-higgsfield-template) (2026-07-29) — Higgsfield AI ad generator on a weekly loop: performance CSV in, fresh briefs and Meta builds out, every ad left paused until you flip it live. Fork on… ([markdown](https://www.jurniti.com/blog/meta-creative-loop-higgsfield-template.md)) - [Linear Triage Agent — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/linear-intake-and-triage-sf-template) (2026-07-29) — Linear triage agent that drafts labels, acceptance criteria, and priority suggestions for unassigned issues. You assign. It never pings your team on its own.… ([markdown](https://www.jurniti.com/blog/linear-intake-and-triage-sf-template.md)) - [AI Lead Scoring: Build a Daily Call List with an Agent](https://www.jurniti.com/blog/lead-triage-call-list-operator-template) (2026-07-29) — AI lead scoring for HubSpot: ranks every lead by fit, engagement, and urgency, then hands you today's top 5 calls with talking points. Read-only. You dial.… ([markdown](https://www.jurniti.com/blog/lead-triage-call-list-operator-template.md)) - [Kilo Code on a MicroVM: The CLI Is Not a Light Install](https://www.jurniti.com/blog/kilocode-on-a-microvm) (2026-07-29) — Kilo Code's CLI ships a ~162MB native binary. Run it on a Firecracker microVM, not your laptop — BYOK, from $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/kilocode-on-a-microvm.md)) - [How to Run an AI Newsletter Generator, Hub-and-Spoke Style](https://www.jurniti.com/blog/hub-and-spoke-content-operator-template) (2026-07-29) — An AI newsletter generator that runs weekly: answer seven questions once, get a full issue plus a week of posts, eight that sell nothing and two that do. Fork… ([markdown](https://www.jurniti.com/blog/hub-and-spoke-content-operator-template.md)) - [How to Run an AI Recruiter: Job Posts + Resume Screening](https://www.jurniti.com/blog/hiring-operator-template) (2026-07-29) — AI recruiter for small-business hiring: drafts a job post in your voice, then ranks pasted resumes on skills only, reasoning shown. It drafts; you hire. Fork… ([markdown](https://www.jurniti.com/blog/hiring-operator-template.md)) - [AI Cofounder Stack — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/hermes-ai-cofounder-stack-template) (2026-07-29) — An AI cofounder with its own inbox, phone number, and debit card. Every send and every charge waits for your yes. Fork it on Hermes, bring your keys. Fork on… ([markdown](https://www.jurniti.com/blog/hermes-ai-cofounder-stack-template.md)) - [How to Run an AI Code Review Agent on Your GitHub PRs](https://www.jurniti.com/blog/github-pr-review-gate-sf-template) (2026-07-29) — AI code review agent for GitHub PRs: risk call-outs, missing tests, a paste-ready review draft. You post, you merge. Nothing auto-merges, ever. Fork on… ([markdown](https://www.jurniti.com/blog/github-pr-review-gate-sf-template.md)) - [GitHub Copilot CLI on a jurniti MicroVM](https://www.jurniti.com/blog/github-copilot-cli-on-a-microvm) (2026-07-29) — Run GitHub Copilot CLI on jurniti: Firecracker microVM, BYOK, paid plans, 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/github-copilot-cli-on-a-microvm.md)) - [Gemini CLI on a jurniti MicroVM](https://www.jurniti.com/blog/gemini-cli-on-a-microvm) (2026-07-29) — Run Gemini CLI on jurniti: Firecracker microVM, BYOK, paid plans, 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/gemini-cli-on-a-microvm.md)) - [Founder Growth Operator — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/founder-growth-operator-for-claude-code-template) (2026-07-29) — Wake up to a written growth plan: traffic notes, SEO ideas, outbound angles, one experiment. Files live on your microVM. Drafts only, never auto-sends. Fork… ([markdown](https://www.jurniti.com/blog/founder-growth-operator-for-claude-code-template.md)) - [Flaky Test Hunter — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/flaky-test-hunter-sf-template) (2026-07-29) — Flaky test hunter for GitHub Actions: groups CI failures by test, scores flake vs real bug, drafts the quarantine or fix PR. You merge every change. Fork on… ([markdown](https://www.jurniti.com/blog/flaky-test-hunter-sf-template.md)) - [Etsy Listing Optimization: Fix Your Title and All 13 Tags](https://www.jurniti.com/blog/etsy-seo-optimizer-template) (2026-07-29) — Paste a listing or drop your shop CSV: 15-25 buyer-language keywords, a retitled draft, and all 13 tags — before/after, drafts only. Nothing touches your shop. ([markdown](https://www.jurniti.com/blog/etsy-seo-optimizer-template.md)) - [Automate Directory Submissions & Backlinks with an Agent](https://www.jurniti.com/blog/directory-campaign-operator-openclaw-template) (2026-07-29) — A directory submission agent that builds your backlink campaign: hand-picked listings, tier-specific copy, a tracker you own. No spam blasts, no agency. Fork… ([markdown](https://www.jurniti.com/blog/directory-campaign-operator-openclaw-template.md)) - [How to Run an AI Customer Service Agent That Only Drafts](https://www.jurniti.com/blog/customer-care-operator-template) (2026-07-29) — AI customer service agent that pulls the customer's history, drafts a reply that names the real problem, and flags the fix. Read-only. It drafts; you send.… ([markdown](https://www.jurniti.com/blog/customer-care-operator-template.md)) - [Higgsfield AI Video Ad Factory — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/client-brand-ad-factory-higgsfield-template) (2026-07-29) — Higgsfield AI video ads from a product URL: UGC, unboxing, and showcase modes with your client's brand rules built in. You review before anything ships. Fork… ([markdown](https://www.jurniti.com/blog/client-brand-ad-factory-higgsfield-template.md)) - [Amp Code on a jurniti MicroVM](https://www.jurniti.com/blog/amp-code-on-a-microvm) (2026-07-29) — Run Amp on jurniti: Firecracker microVM, BYOK, paid plans, 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/amp-code-on-a-microvm.md)) - [Run Aider on a Remote MicroVM (Not Just Your Laptop)](https://www.jurniti.com/blog/aider-on-a-microvm) (2026-07-29) — Run Aider on jurniti: Firecracker microVM, BYOK, paid plans, 30-day money-back. No free trial. ([markdown](https://www.jurniti.com/blog/aider-on-a-microvm.md)) - [AI Vulnerability Scanner on Claude Code: CodeCrucible SAST](https://www.jurniti.com/blog/ai-vulnerability-scanner-codecrucible-template) (2026-07-29) — Run an AI SAST / AI vulnerability scanner on a Claude Code microVM: Block CodeCrucible, dry-run cost gate, SARIF out. You verify findings. No free trial. ([markdown](https://www.jurniti.com/blog/ai-vulnerability-scanner-codecrucible-template.md)) - [AI Automation Agency Template: Run the $999 Audit Yourself](https://www.jurniti.com/blog/ai-automation-agency-template) (2026-07-29) — Fork a Hermes microVM that prospects, teases free mini-audits, sells a $999 AI audit on your Stripe, and drafts deliverables you approve. No free trial. ([markdown](https://www.jurniti.com/blog/ai-automation-agency-template.md)) - [Agent Claim vs Diff Gate — Run It on a jurniti MicroVM](https://www.jurniti.com/blog/agent-claim-vs-diff-gate-sf-template) (2026-07-29) — Code review gate for AI agent PRs: compares the PR body to the real git diff and flags hallucinated claims like 'tests pass'. You keep the merge button. Fork… ([markdown](https://www.jurniti.com/blog/agent-claim-vs-diff-gate-sf-template.md)) - [What Is a Forward Deployed Engineer?](https://www.jurniti.com/blog/what-is-a-forward-deployed-engineer) (2026-07-28) — A forward deployed engineer is a customer-facing software engineer who ships outcomes on a platform—not a ticket-taker and not a pure back-office SWE. ([markdown](https://www.jurniti.com/blog/what-is-a-forward-deployed-engineer.md)) - [Software Factory Agent Templates (Humans Still Merge)](https://www.jurniti.com/blog/software-factory-agent-templates) (2026-07-28) — PR review, flakes, Sentry drafts, Linear triage — agent templates for eng teams. Humans still read and merge. jurniti Software Factory collection. ([markdown](https://www.jurniti.com/blog/software-factory-agent-templates.md)) - [Small Business AI Agent Templates (You Still Approve)](https://www.jurniti.com/blog/small-business-ai-agent-templates) (2026-07-28) — QuickBooks, Stripe, Square, HubSpot — agent templates for owners. Drafts in a microVM; you review and send. jurniti Small Business collection. ([markdown](https://www.jurniti.com/blog/small-business-ai-agent-templates.md)) - [Self-Host mem0 for AI Agents on a MicroVM](https://www.jurniti.com/blog/self-host-mem0-for-ai-agents) (2026-07-28) — mem0 as a jurniti memory plugin: dedicated microVM, grant to agents, BYOK. Low-volume keyword, real isolation. 30-day money-back, from $25/mo. ([markdown](https://www.jurniti.com/blog/self-host-mem0-for-ai-agents.md)) - [Self-Host Langfuse for AI Agents on a MicroVM](https://www.jurniti.com/blog/self-host-langfuse-for-ai-agents) (2026-07-28) — Langfuse for agent traces on jurniti: a dedicated observability microVM, grant to your agents, BYOK. Not a shared SaaS-only story. 30-day money-back. ([markdown](https://www.jurniti.com/blog/self-host-langfuse-for-ai-agents.md)) - [Palantir Forward Deployed Engineer: What the Model Taught](https://www.jurniti.com/blog/palantir-forward-deployed-engineer) (2026-07-28) — Palantir popularized FDE as design partnership at enterprise scale. The durable lesson is platform primitives—not 55 custom repos or pure body shops. ([markdown](https://www.jurniti.com/blog/palantir-forward-deployed-engineer.md)) - [OpenHands on a MicroVM: Cloud Runtime Without Shared Kernels](https://www.jurniti.com/blog/openhands-on-a-microvm) (2026-07-28) — OpenHands cloud demand is real. jurniti runs OpenHands on a Firecracker tenant microVM — BYOK, always-on, from $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/openhands-on-a-microvm.md)) - [OpenClaw Security: Protect the Agent, Not Just the Model](https://www.jurniti.com/blog/openclaw-security) (2026-07-28) — OpenClaw security means isolating the bot that holds chat bridges and keys. Firecracker per tenant, BYOK, no anti-harness FUD. 30-day money-back. ([markdown](https://www.jurniti.com/blog/openclaw-security.md)) - [OpenAI Forward Deployed Engineer: Closing Demo-to-Production](https://www.jurniti.com/blog/openai-forward-deployed-engineer) (2026-07-28) — Model labs hire FDEs to operationalize AI. The parallel for agent platforms: someone must close the gap between demo and production. ([markdown](https://www.jurniti.com/blog/openai-forward-deployed-engineer.md)) - [MCP Security: Isolate Tools Before You Trust Them](https://www.jurniti.com/blog/mcp-security) (2026-07-28) — MCP security is a runtime problem: tool calls can touch shells and secrets. Firecracker isolation + least privilege. BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/mcp-security.md)) - [How to Become a Forward Deployed Engineer](https://www.jurniti.com/blog/how-to-become-a-forward-deployed-engineer) (2026-07-28) — A practical path into FDE: engineering bar, customer surface, platform discipline, and agent deploy craft—without cargo-culting the title. ([markdown](https://www.jurniti.com/blog/how-to-become-a-forward-deployed-engineer.md)) - [GTM Agent Templates That Never Auto-Send](https://www.jurniti.com/blog/gtm-agent-templates-that-never-auto-send) (2026-07-28) — Clay waterfalls, Apollo drafts, LinkedIn openers — in their own microVMs. Your credits. You hit send. jurniti GTM collection, 30-day money-back. ([markdown](https://www.jurniti.com/blog/gtm-agent-templates-that-never-auto-send.md)) - [Forward Deployed Engineer vs Software Engineer](https://www.jurniti.com/blog/forward-deployed-engineer-vs-software-engineer) (2026-07-28) — Same engineering bar; different surface. FDEs own customer outcomes on a platform. Pure product SWE may never sit with the buyer. ([markdown](https://www.jurniti.com/blog/forward-deployed-engineer-vs-software-engineer.md)) - [Forward Deployed Engineer Salary: How to Read the Ranges](https://www.jurniti.com/blog/forward-deployed-engineer-salary) (2026-07-28) — FDE pay reflects customer-facing engineering risk and delivery pressure. Use ranges as signal, then judge the work: outcomes, isolation, fleet ops. ([markdown](https://www.jurniti.com/blog/forward-deployed-engineer-salary.md)) - [Forward Deployed Engineer Job: What the Role Actually Is](https://www.jurniti.com/blog/forward-deployed-engineer-job) (2026-07-28) — FDE jobs sit between product engineering and customer delivery. What hiring teams mean, day-one work, and the deploy craft agent FDEs need. ([markdown](https://www.jurniti.com/blog/forward-deployed-engineer-job.md)) - [Forward Deployed AI Engineer: When the Product Is an Agent](https://www.jurniti.com/blog/forward-deployed-ai-engineer) (2026-07-28) — An AI-native FDE ships agent outcomes with isolation, BYOK, and always-on runtimes—not laptop demos that die at the airport. ([markdown](https://www.jurniti.com/blog/forward-deployed-ai-engineer.md)) - [Cline on a MicroVM: Beyond the Editor Sidebar](https://www.jurniti.com/blog/cline-on-a-microvm) (2026-07-28) — Cline cloud searchers want more than a VS Code sidebar. Always-on Cline in a Firecracker guest, BYOK, from $25/mo with 30-day money-back. ([markdown](https://www.jurniti.com/blog/cline-on-a-microvm.md)) - [Claude Code Worktrees: Multiple Agents Done Right](https://www.jurniti.com/blog/claude-code-worktrees) (2026-07-28) — Claude Code worktrees and multiple agents on one laptop share a kernel and credentials. One microVM per agent is the fix. BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/claude-code-worktrees.md)) - [Claude Code Security: Stop Running Agents on a Shared Kernel](https://www.jurniti.com/blog/claude-code-security) (2026-07-28) — Claude Code security is a runtime problem. Docker shares a kernel. Firecracker microVMs don't. BYOK, 30-day money-back, from $25/mo. ([markdown](https://www.jurniti.com/blog/claude-code-security.md)) - [Anthropic Forward Deployed Engineer: The Lab FDE Signal](https://www.jurniti.com/blog/anthropic-forward-deployed-engineer) (2026-07-28) — Lab FDE hiring means applied AI is won in customer environments. If you ship agents for others, deploy craft is the skill under the title. ([markdown](https://www.jurniti.com/blog/anthropic-forward-deployed-engineer.md)) - [AI Agent Sandbox for Customer Deployments](https://www.jurniti.com/blog/ai-agent-sandbox-for-customer-deployments) (2026-07-28) — Customer-scoped agents need isolation and always-on, not a shared-kernel free-for-all. How FDEs should think about sandboxes, fleets, and production. ([markdown](https://www.jurniti.com/blog/ai-agent-sandbox-for-customer-deployments.md)) - [AI Agent Platform: What Buyers Actually Need](https://www.jurniti.com/blog/ai-agent-platform) (2026-07-28) — An AI agent platform is not a chat UI. Isolation, BYOK, always-on runtime, and harness choice. jurniti microVMs from $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/ai-agent-platform.md)) - [Self-Host Poolside pool on a Server, 24/7](https://www.jurniti.com/blog/pool-hosting) (2026-07-21) — Run Poolside's pool coding agent on its own Firecracker microVM. Laguna-native, ACP-ready, BYOK. 30-day money-back. ([markdown](https://www.jurniti.com/blog/pool-hosting.md)) - [Aider vs Cline vs 4 More CLI Coding Agents](https://www.jurniti.com/blog/cli-coding-agents-on-microvms) (2026-07-20) — Aider vs Cline vs Codex, Claude Code, Goose and OpenCode — runtime, install, auth, and which support headless fleet dispatch. From real adapters. ([markdown](https://www.jurniti.com/blog/cli-coding-agents-on-microvms.md)) - [Self-Hosted vs Managed Claude Code: An Honest Comparison](https://www.jurniti.com/blog/self-hosted-vs-managed-claude-code) (2026-07-17) — Self-host Claude Code on a VPS or pay for managed hosting? A straight comparison of setup, security, persistence, and the real cost of your own time. ([markdown](https://www.jurniti.com/blog/self-hosted-vs-managed-claude-code.md)) - [AI Agent Sandbox: MicroVM vs Docker, Compared](https://www.jurniti.com/blog/microvm-vs-docker-agent-hosting) (2026-07-17) — Choosing an AI agent sandbox? Containers share the host kernel; Firecracker microVMs don't. Why sub-$5 agent hosting can't include a kernel boundary. ([markdown](https://www.jurniti.com/blog/microvm-vs-docker-agent-hosting.md)) - [Self-Host OpenScience on a microVM](https://www.jurniti.com/blog/openscience-hosting) (2026-07-15) — Run OpenScience on a Firecracker microVM — literature, code, experiments, and write-up in one browser workspace. BYOK, persistent home, snapshot/fork. ([markdown](https://www.jurniti.com/blog/openscience-hosting.md)) - [Self-Host Mastra Agents in Production](https://www.jurniti.com/blog/mastra-hosting) (2026-07-05) — Run Mastra, the TypeScript agent framework, on its own Firecracker microVM. Studio at your TLS subdomain, persistent volume. BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/mastra-hosting.md)) - [Self-Host DeepAgents Code on a microVM](https://www.jurniti.com/blog/deepagents-code-hosting) (2026-06-30) — Run LangChain's DeepAgents Code (dcode) on its own Firecracker microVM. Any provider:model via OpenRouter, no frontier-lab subscription. 30-day money-back. ([markdown](https://www.jurniti.com/blog/deepagents-code-hosting.md)) - [Self-Host Paperclip on a Server](https://www.jurniti.com/blog/paperclip-hosting) (2026-06-07) — Run Paperclip, the open-source manager for a team of AI agents, on its own Firecracker microVM. Dashboard at your TLS subdomain. BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/paperclip-hosting.md)) - [Self-Host OpenCode: BYOK Any Model, No Markup](https://www.jurniti.com/blog/opencode-hosting) (2026-06-07) — Self-host OpenCode 24/7. BYOK any model — Claude, GPT, Gemini, Ollama — no markup. Persistent home on a Firecracker microVM. From $25/mo. ([markdown](https://www.jurniti.com/blog/opencode-hosting.md)) - [Self-Host Grok Build: Headless, Parallel Subagents](https://www.jurniti.com/blog/grok-build-hosting) (2026-06-07) — Run xAI's Grok Build, with parallel subagents and Plan Mode, on its own Firecracker microVM. BYO Grok subscription, persistent home. 30-day money-back. ([markdown](https://www.jurniti.com/blog/grok-build-hosting.md)) - [Self-Host Goose on a Server, 24/7](https://www.jurniti.com/blog/goose-ai-agent-hosting) (2026-06-07) — Run Goose, the open AI agent stewarded by the Linux Foundation, on its own Firecracker microVM. 15+ providers, 70+ MCP extensions. BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/goose-ai-agent-hosting.md)) - [Self-Host Codex CLI: ChatGPT Login, Always On](https://www.jurniti.com/blog/codex-cli-hosting) (2026-06-07) — Self-host OpenAI Codex CLI 24/7. Sign in with your ChatGPT plan, keep AGENTS.md — Firecracker microVM instead of a shared VPS. From $25/mo. ([markdown](https://www.jurniti.com/blog/codex-cli-hosting.md)) - [Run Claude Code on a Server (Remote microVM)](https://www.jurniti.com/blog/claude-code-hosting) (2026-06-07) — Run Claude Code on a server: always-on Firecracker microVM, your Claude login or API key stays in the guest, not a shared container. 30-day money-back. ([markdown](https://www.jurniti.com/blog/claude-code-hosting.md)) - [Self-Host the Pi Coding Agent on a Server](https://www.jurniti.com/blog/pi-coding-agent-hosting) (2026-06-02) — Run Mario Zechner's Pi coding agent on its own Firecracker microVM. BYO Anthropic or OpenAI key, persistent workspace, optional web TUI. 30-day money-back. ([markdown](https://www.jurniti.com/blog/pi-coding-agent-hosting.md)) - [Self-Host OpenClaw 24/7: VPS vs Its Own microVM](https://www.jurniti.com/blog/openclaw-hosting) (2026-06-02) — Self-host OpenClaw 24/7 on a dedicated Firecracker microVM instead of a shared-kernel VPS. WhatsApp, Telegram, Discord, Signal bridges. BYOK, 30-day money-back. ([markdown](https://www.jurniti.com/blog/openclaw-hosting.md)) - [Self-Host Hermes Agent: Always-On microVM Guide](https://www.jurniti.com/blog/hermes-agent-hosting) (2026-06-02) — Self-host the Nous Research Hermes Agent on a Firecracker microVM — BYOK, persistent home, messaging bridges. From $25/mo, 30-day money-back. ([markdown](https://www.jurniti.com/blog/hermes-agent-hosting.md)) --- ## Public templates Forkable snapshots published by the community (125 public). Open any page and use the Fork action to provision a new VM seeded with it. - [Clay Outbound Engine](https://www.jurniti.com/templates/clay-outbound-engine) (Claude Code, 0 forks, 0 stars) — Find 20 CEOs. Get work emails. Draft openers. Clay’s search→enrich loop, preinstalled on Claude Code — fork it, clay login, go. Your credits. - [First Customer Finder for Codex](https://www.jurniti.com/templates/first-customer-finder-for-codex) (Codex CLI, 0 forks, 0 stars) — Codex finds evidence-backed first-customer prospects: ICP, scored shortlist, source links, openers, HTML report. Never auto-sends. BYOK. - [Hermes Full GTM OS](https://www.jurniti.com/templates/hermes-full-gtm-os) (Hermes Agent, 0 forks, 0 stars) — Hermes GTM OS: ICP → traffic → capture → nurture → qualify → convert → retain. File-native tables, signal-led outbound, human-gated sends. - [Marketing Skills for Claude Code](https://www.jurniti.com/templates/marketing-skills-for-claude-code) (Claude Code, 0 forks, 0 stars) — Corey Haines’ Marketing Skills (v2.6.0, 46 skills, MIT) on Claude Code: CRO, copy, SEO, ads, email, pricing, launch — plus marketing-loops. - [On-Demand Leads with Apify](https://www.jurniti.com/templates/on-demand-leads-with-apify) (Claude Code, 0 forks, 0 stars) — Stop buying static lead lists. Claude Code + Apify MCP builds on-demand lists from a URL — ICP, scrape, email fix, engagers. Never auto-sends. - [Outbound Loop for Codex](https://www.jurniti.com/templates/outbound-loop-for-codex) (Codex CLI, 0 forks, 0 stars) — Outbound as a dual loop on Codex: daily score→draft→log, weekly one-concept tune behind an eval gate. Never auto-sends. Your sequencer. - [Higgsfield AI Meta Ad Creative Loop](https://www.jurniti.com/templates/meta-creative-loop-higgsfield) (Claude Code, 0 forks, 0 stars) — Higgsfield AI ad generator on a weekly loop: performance CSV in, fresh briefs and Meta builds out, every ad left paused until you flip it live. - [AI Lead Scoring + Call List Operator](https://www.jurniti.com/templates/lead-triage-call-list-operator) (Hermes Agent, 0 forks, 0 stars) — AI lead scoring for HubSpot: ranks every lead by fit, engagement, and urgency, then hands you today's top 5 calls with talking points. Read-only. You dial. - [Monday Brief Operator](https://www.jurniti.com/templates/monday-brief-operator) (Hermes Agent, 0 forks, 0 stars) — QuickBooks + HubSpot in one Monday brief: cash on hand, stuck deals, and your top 3 moves before your coffee's cold. Read-only - it reports, you act. - [AI SEO Agent - BOFU Content Engine](https://www.jurniti.com/templates/saas-bofu-seo-content-engine-for-claude-code) (Claude Code, 0 forks, 0 stars) — An AI SEO agent that drafts bottom-funnel comparison pages with CTAs built in. Checks page one first. Never auto-publishes. Fork it, boot in about 3 minutes. - [Founder Growth Operator](https://www.jurniti.com/templates/founder-growth-operator-for-claude-code) (Claude Code, 0 forks, 0 stars) — Wake up to a written growth plan: traffic notes, SEO ideas, outbound angles, one experiment. Files live on your microVM. Drafts only, never auto-sends. - [Sentry AI Agent: Error to Draft PR](https://www.jurniti.com/templates/sentry-error-to-draft-pr-sf) (Claude Code, 0 forks, 0 stars) — Sentry AI agent that turns a production error into a draft fix PR: stack summary, suspect commits, minimal patch. You review and merge. Nothing auto-ships. - [Linear Triage Agent](https://www.jurniti.com/templates/linear-intake-and-triage-sf) (Claude Code, 0 forks, 0 stars) — Linear triage agent that drafts labels, acceptance criteria, and priority suggestions for unassigned issues. You assign. It never pings your team on its own. - [AI Code Review Agent for GitHub PRs](https://www.jurniti.com/templates/github-pr-review-gate-sf) (Claude Code, 0 forks, 0 stars) — AI code review agent for GitHub PRs: risk call-outs, missing tests, a paste-ready review draft. You post, you merge. Nothing auto-merges, ever. - [Flaky Test Hunter](https://www.jurniti.com/templates/flaky-test-hunter-sf) (Claude Code, 0 forks, 0 stars) — Flaky test hunter for GitHub Actions: groups CI failures by test, scores flake vs real bug, drafts the quarantine or fix PR. You merge every change. - [Agent Claim vs Diff Gate](https://www.jurniti.com/templates/agent-claim-vs-diff-gate-sf) (Claude Code, 0 forks, 0 stars) — Code review gate for AI agent PRs: compares the PR body to the real git diff and flags hallucinated claims like 'tests pass'. You keep the merge button. - [Nightly One-Antipattern Fixer](https://www.jurniti.com/templates/nightly-one-antipattern-fixer-sf) (Claude Code, 0 forks, 0 stars) — Overnight code cleanup agent: finds one safe lint or type fix per night and opens one small draft PR. Never touches auth, billing, or migrations. - [Sentry AI Skills Pack for Claude Code](https://www.jurniti.com/templates/sentry-for-ai-skills-pack-sf) (Claude Code, 0 forks, 0 stars) — Sentry AI skills on your own isolated Claude Code VM. First-party getsentry skills, your token, you approve every monitor change. 30-day money-back. - [Etsy SEO Optimizer](https://www.jurniti.com/templates/etsy-seo-optimizer) (Claude Code, 0 forks, 0 stars) — Turn the listings you already have into search traffic. Paste one (or drop your shop CSV) and get keyword research, a retitled draft, and all 13 tags — as drafts you apply yourself. Nothing ever touches your shop. - [OpenAI Codex Security for Codex](https://www.jurniti.com/templates/openai-codex-security-for-codex) (Codex CLI, 0 forks, 0 stars) — Fork a Codex microVM with OpenAI's security CLI + skill pack installed. Dry-run first. Your keys. Your findings. - [Everything Claude Code: 64 Agents](https://www.jurniti.com/templates/ecc) (Claude Code, 0 forks, 0 stars) — Fork affaan’s grand-prize Everything Claude Code — 64 agents, 261 skills, hooks & rules — on your microVM. Your Claude login; live in ~3 min. - [AI Automation Agency](https://www.jurniti.com/templates/ai-automation-agency) (Hermes Agent, 0 forks, 0 stars) — Run a one-person AI automation agency: this Hermes agent prospects, sends teaser AI audits, and bills through your own Stripe. Bring an OpenRouter key. - [AI Drug Discovery Agent (UniProt/PDB)](https://www.jurniti.com/templates/comp-bio-research-agent-for-openscience) (OpenScience, 0 forks, 0 stars) — Target briefs for AI drug discovery, grounded in UniProt, PDB and ChEMBL accessions you can click and check. Bring one model key. Nothing docked or predicted. - [Landing Page Copy Audit (Halbert)](https://www.jurniti.com/templates/copy-url-monthly-audit-openclaw) (OpenClaw, 0 forks, 0 stars) — Audit one landing page URL against Halbert's 17 sales rules and Ogilvy's 10 writing rules, monthly on cron. Markdown report file. Bring an OpenClaw model key. - [GLM-5.2 Coding Agent (OpenRouter)](https://www.jurniti.com/templates/deepagents-code-glm-5-2-on-openrouter) (DeepAgents Code, 0 forks, 0 stars) — DeepAgents Code pre-wired to GLM-5.2 over OpenRouter with the shell tool on. Read, edit, run — no model routing to configure. Bring one OpenRouter key. - [Directory Submission Service Agent](https://www.jurniti.com/templates/directory-campaign-operator-openclaw) (OpenClaw, 0 forks, 0 stars) — A directory submission service you run yourself: fit-checked listings, per-site copy, a tracker you own. OpenClaw, bring your own model key. - [AI Cofounder Stack](https://www.jurniti.com/templates/hermes-ai-cofounder-stack) (Hermes Agent, 0 forks, 0 stars) — An AI cofounder with its own inbox, phone number and spend card. Every send and every charge waits for your yes. Hermes, bring your own keys. - [Newsjacking Agent: PR Angles](https://www.jurniti.com/templates/hermes-newsjack-agentic-pr-team) (Hermes Agent, 0 forks, 0 stars) — A newsjacking agent with 20 PR skills: story angles, headlines, fact-check, journalist fit, coverage tracking. Hermes, bring your own model key. - [AI Newsletter Generator](https://www.jurniti.com/templates/hub-and-spoke-content-operator) (Hermes Agent, 0 forks, 0 stars) — An AI newsletter generator that runs weekly: one long issue, six to ten posts, two that sell. Nothing publishes without your yes. Bring an OpenRouter key. - [Swipe File Copywriter](https://www.jurniti.com/templates/hermes-swipe-file-copywriter) (Hermes Agent, 0 forks, 0 stars) — A copywriter that searches a private swipe file before it writes: landing pages, ads, emails, cold openers, plus a scored critique. Bring an OpenRouter key. - [Promo Video Maker (HyperFrames)](https://www.jurniti.com/templates/hyperframes-promo-video-studio) (Claude Code, 0 forks, 0 stars) — Describe the launch, get a rendered promo video back. Claude Code with 17 HyperFrames skills installed. Bring your Claude login and a HyperFrames key. - [Immersive Reader (Codex)](https://www.jurniti.com/templates/immersive-reader-codex) (Codex CLI, 0 forks, 0 stars) — Turn one long essay, paper, or transcript into an interactive reading site: chapters, search, highlights, notes, bilingual, light/dark. Bring an OpenAI key. - [Listmonk Newsletter Agent](https://www.jurniti.com/templates/listmonk-newsletter-agent) (Claude Code, 0 forks, 0 stars) — Claude Code agent that drafts the issue and drives self-hosted listmonk over its admin API: draft, test-send, schedule. Bring your listmonk token. - [Local SEO Agency Page Builder](https://www.jurniti.com/templates/local-landing-page-agency) (Hermes Agent, 0 forks, 0 stars) — Build a branded Astro landing page for a local business, email the live preview, charge with Stripe, go live on their domain. Hermes + your OpenRouter key. - [AI Co-Scientist for ML Research](https://www.jurniti.com/templates/ml-co-scientist-for-openscience) (OpenScience, 0 forks, 0 stars) — OpenScience co-scientist for ML: literature loop, one concrete experiment, train/eval scaffold, methods draft — as files. Bring your own Anthropic key. - [vLLM Inference Lab: p99 Bench](https://www.jurniti.com/templates/ml-inference-lab-for-openscience) (OpenScience, 0 forks, 0 stars) — OpenScience lab for OpenAI-compatible inference: design a vLLM-style serve stack, bench it, read real p50/p99. BYOK Anthropic plus optional HF token. - [Apollo Alternative: Pay-Per-Lead](https://www.jurniti.com/templates/monid-apollo-pay-per-lead) (Claude Code, 0 forks, 0 stars) — Apollo alternative for Claude Code: pull B2B contacts with verified work emails at about $0.05 per enriched lead. Bring a Monid key. Never auto-sends. - [Whiteboard Animation Ads, Hand-Drawn](https://www.jurniti.com/templates/notebook-doodle-ad-agent) (Claude Code, 0 forks, 0 stars) — Hand-drawn vertical ads on ruled paper, pencil inked where your product traces it. One approval, then paste-ready Higgsfield and Seedance prompts. - [Keyword Research Agent (OpenSEO)](https://www.jurniti.com/templates/openseo-agent-claude-code) (Claude Code, 0 forks, 0 stars) — Keyword research in Claude Code that reads live data: 8 SEO skills plus OpenSEO MCP for volume, SERPs, backlinks and GSC. Bring a Claude login. - [Postiz Social Media Scheduler](https://www.jurniti.com/templates/postiz-social-media-scheduler) (OpenClaw, 0 forks, 0 stars) — Postiz + OpenClaw: tell one agent what to post and when. It drafts and queues across the 28 channels Postiz lists. Bring a Postiz API key. - [Smartlead AI Cold Email Operator](https://www.jurniti.com/templates/smartlead-cold-email-operator) (Claude Code, 0 forks, 0 stars) — Smartlead AI on Claude Code: paste your list, get a draft campaign, an opener per lead, a warm-inbox check. Nothing sends. Bring a Smartlead API key. - [Stripe AI Churn Analysis Report](https://www.jurniti.com/templates/stripe-ai-churn-analysis-revenue-report) (Claude Code, 0 forks, 0 stars) — Weekly churn analysis and revenue report from Stripe, posted to Slack. Bring a read-only Stripe key; a grounding check blocks any figure it can't source. - [AI Tutor: Adaptive Learning Agent](https://www.jurniti.com/templates/teacher-adaptive-learn-tutor-codex) (Codex CLI, 0 forks, 0 stars) — An AI tutor on Codex that teaches any topic at your level, then writes an interactive HTML lesson you keep. Worked example and quiz. Bring an OpenAI key. - [Marketing AI Agent: X Launch Operator](https://www.jurniti.com/templates/viral-x-launch-operator-marketing-skills) (Claude Code, 0 forks, 0 stars) — A marketing AI agent that runs X launch day: builds the kit, posts to X through Postiz, drafts warm pings, works the first hour. Bring a Postiz API key. - [QuickBooks AI Cash Flow Operator](https://www.jurniti.com/templates/quickbooks-cash-flow-invoice-operator) (Hermes Agent, 0 forks, 0 stars) — QuickBooks AI for 'will I make payroll?' - 30/60/90-day cash forecast from your real books + drafted overdue-invoice reminders. It drafts; you send. - [AI Video Ads: Vox Paper-Collage Style](https://www.jurniti.com/templates/vox-style-paper-collage-ad) (Claude Code, 0 forks, 0 stars) — AI video ads in the Vox paper-collage look. Brief it 30 seconds, approve once, get chained Omni Flash clips or paste-ready prompts. Bring your model key. - [AI Recruiter: Resume Screening](https://www.jurniti.com/templates/hiring-operator) (Hermes Agent, 0 forks, 0 stars) — AI recruiter for small-business hiring: drafts a job post in your voice, then ranks pasted resumes on skills only, reasoning shown. It drafts; you hire. - [AI Customer Service Agent](https://www.jurniti.com/templates/customer-care-operator) (Hermes Agent, 0 forks, 0 stars) — AI customer service agent that pulls the customer's history, drafts a reply that names the real problem, and flags the fix. Read-only. It drafts; you send. - [CodeCrucible AI Vulnerability Scanner](https://www.jurniti.com/templates/ai-vulnerability-scanner-codecrucible-sast) (Claude Code, 0 forks, 0 stars) — AI vulnerability scanner / AI SAST on Claude Code: Block CodeCrucible, dry-run cost gate, SARIF out. You verify findings — nothing auto-merges. - [AI Ad Generator: Paid Acquisition](https://www.jurniti.com/templates/paid-acquisition-agent-jockey-for-claude-code) (Claude Code, 0 forks, 0 stars) — AI ad generator that reads its own results: research pains, stack angles, batch creatives, log winners. Nothing spends without you. From $25/mo. - [Founder Playbook Skills](https://www.jurniti.com/templates/founder-playbook-skills-for-claude-code) (Claude Code, 0 forks, 0 stars) — 30 people loved it. Zero paid. Claude Code with 15 founder skills — diagnose picks one book and one concrete move this week. BYOK. - [Reddit Marketing Tool: Ad Angles](https://www.jurniti.com/templates/mine-reddit-for-the-ad-angle-for-claude-code) (Claude Code, 0 forks, 0 stars) — Reddit marketing tool for Claude Code: mine real complaints, rank the top 3 pains, write ads in your customers' words. Nothing spends without you. - [Ads Dashboard + ROAS Board](https://www.jurniti.com/templates/paid-ads-roas-board-for-hermes) (Hermes Agent, 0 forks, 0 stars) — Ads dashboard with a budget gate: drafts hooks and CTAs to an Obsidian board you own, sets break-even ROAS before a dollar of spend. One channel, mastered. - [rpi-qrsp-compaction-pack-sf](https://www.jurniti.com/templates/rpi-qrsp-compaction-pack-sf) (Claude Code, 0 forks, 0 stars) — Research → plan → implement skills with intentional compaction. Human gates on research and plan. Draft only — you still merge. - [program-design-gate-sf](https://www.jurniti.com/templates/program-design-gate-sf) (Claude Code, 0 forks, 0 stars) — Call-stack trees, file-tree diffs, and types before implementation. Architecture alone is not enough. You approve design first. - [diff-tree-pr-annotator-sf](https://www.jurniti.com/templates/diff-tree-pr-annotator-sf) (Claude Code, 0 forks, 0 stars) — Large PRs get a logical diff tree — ordered by change intent, not alpha. Review energy heuristic. You still read the code. - [product-review-mockup-sf](https://www.jurniti.com/templates/product-review-mockup-sf) (Claude Code, 0 forks, 0 stars) — Problem, success metric, and rough HTML mockups before any code. Mockups are cheaper than slop PRs. - [vertical-slice-planner-sf](https://www.jurniti.com/templates/vertical-slice-planner-sf) (Claude Code, 0 forks, 0 stars) — Rewrite horizontal migrations→API→UI plans into tracer-bullet slices you can curl and review every 100–200 lines. - [maintainability-slop-meter-sf](https://www.jurniti.com/templates/maintainability-slop-meter-sf) (Claude Code, 0 forks, 0 stars) — Directional maintainability smells on a PR — complexity, duplication, shotgun risk. Draft review only. Tests green ≠ maintainable. - [architecture-review-gate-sf](https://www.jurniti.com/templates/architecture-review-gate-sf) (Claude Code, 0 forks, 0 stars) — Sequence, contracts, and data model deltas before sprawl. Lights on for long-lived decisions. Pair with program design. - [ci-green-keeper-sf](https://www.jurniti.com/templates/ci-green-keeper-sf) (Claude Code, 0 forks, 0 stars) — When main goes red, draft the smallest green fix PR. Human merge gate. Never force-push main. - [github-migration-review-sf](https://www.jurniti.com/templates/github-migration-review-sf) (Claude Code, 0 forks, 0 stars) — Schema/migration diffs with expand/contract safety, lock risk, and rollback notes. You still ship. - [sentry-incident-brief-sf](https://www.jurniti.com/templates/sentry-incident-brief-sf) (Claude Code, 0 forks, 0 stars) — Error cluster → timeline, suspect commits, draft postmortem. You page. Never auto-mute or auto-deploy. - [Paid Ads Deanon Instantly Plunk Drip for Hermes](https://www.jurniti.com/templates/paid-ads-deanon-instantly-plunk-drip-for-hermes) (Hermes Agent, 0 forks, 0 stars) — Cheat-code demand loop: landing + ads (Higgsfield), de-anon, ICP, MillionVerifier, Instantly drafts, Plunk nurture drips, inbox demos. RB2B optional. - [Loopany Reddit Karma for Claude Code](https://www.jurniti.com/templates/loopany-reddit-karma-for-claude-code) (Claude Code, 0 forks, 0 stars) — Fork a Claude Code box with the Loopany Reddit karma loop: wiki, filters, ledger, OpenCLI. Draft-first. Your account. - [Gorgias Alternative: Support Desk](https://www.jurniti.com/templates/gorgias-alternative-support-desk) (Claude Code, 0 forks, 0 stars) — Draft Shopify support replies and review responses on your own read-only token — no per-resolution meter. It writes drafts. You send them. - [Pi Workflows Control Flow](https://www.jurniti.com/templates/pi-workflows-control-flow) (Pi, 0 forks, 0 stars) — Control flow inside Pi: TypeScript graphs with y/n branches, implement-verify loops, seven seeded workflows, live graph widget. Bring your model key. - [Secret Scan Triage](https://www.jurniti.com/templates/secret-scan-triage) (Claude Code, 0 forks, 0 stars) — Triages GitHub secret-scanning alerts by blast radius into a rotate-first runbook. Never closes an alert, never prints the secret. - [Dependabot PR Triage](https://www.jurniti.com/templates/dependabot-pr-triage) (Claude Code, 0 forks, 0 stars) — Sorts a Dependabot backlog into routine, needs-migration and security — then drafts the groups and ignore rules that stop it regenerating. Never merges. - [Datadog Alert Noise Digest](https://www.jurniti.com/templates/datadog-alert-noise-digest) (Claude Code, 0 forks, 0 stars) — Ranks Datadog monitors by action rate, not volume — then proposes specific thresholds with the percentile behind them. Never mutes a monitor. - [Month-End Close Prep](https://www.jurniti.com/templates/month-end-close-prep) (Claude Code, 0 forks, 0 stars) — Reconciles bank, processor, AR and AP exports into a close packet with tie-outs shown. Unexplained variance is a finding — it never plugs. Never posts entries. - [Inventory Reorder Forecast](https://www.jurniti.com/templates/inventory-reorder-forecast) (Claude Code, 0 forks, 0 stars) — Days-of-cover reorder brief from your sales and stock exports — and it flags the SKUs whose velocity a stockout quietly suppressed. Never orders. - [PayPal Dispute Response Drafter](https://www.jurniti.com/templates/paypal-dispute-response-drafter) (Claude Code, 0 forks, 0 stars) — Drafts PayPal dispute responses sorted by deadline, with every claim matched to an artifact and the unprovable ones listed as gaps. Never submits. - [Margin & Pricing Check](https://www.jurniti.com/templates/margin-pricing-check) (Claude Code, 0 forks, 0 stars) — True margin from landed cost — not markup, not invoice cost — plus price scenarios with the break-even volume change. Never changes prices. - [Engineering Docs Rot Finder](https://www.jurniti.com/templates/engineering-docs-rot-finder) (Claude Code, 0 forks, 0 stars) — Finds the docs your codebase contradicts — dead commands, moved paths, broken links — with sourced corrections. Age isn't rot, and it never edits your wiki. - [Amazon PPC Analyst](https://www.jurniti.com/templates/amazon-ppc-analyst) (Claude Code, 0 forks, 0 stars) — Finds the Sponsored Products spend that never converted, then drafts harvest, negate and bid-down changes with the arithmetic shown. Never changes bids. - [eBay Listing Drafter](https://www.jurniti.com/templates/ebay-listing-drafter) (Claude Code, 0 forks, 0 stars) — Drafts eBay listings that fit the 80-char title cap, fill every item specific, and price from sold comps — not asking prices. Never lists. - [Linear Bug Repro Pack](https://www.jurniti.com/templates/linear-bug-repro-pack) (Claude Code, 0 forks, 0 stars) — Tests every Linear bug against the four things that make it reproducible, then drafts the exact questions for whatever is missing. Never closes an issue. - [Sales Battle Card Generator](https://www.jurniti.com/templates/sales-battle-card-generator) (Claude Code, 0 forks, 0 stars) — Internal battle cards where every competitor claim is dated and sourced — and the 'when they win' section is mandatory. Never publishes externally. - [Slack Daily Brief](https://www.jurniti.com/templates/slack-daily-brief) (Claude Code, 0 forks, 0 stars) — A five-line Slack brief that leads with what changed — and says so in one line when nothing did. You approve every post; it never schedules. - [Waterfall Enrichment Cost Auditor](https://www.jurniti.com/templates/waterfall-enrichment-cost-auditor) (Claude Code, 0 forks, 0 stars) — Reorders your enrichment waterfall by cost-per-incremental-hit and finds the providers billing for rows already filled. Never spends a credit. - [Support Ticket Deflector](https://www.jurniti.com/templates/support-ticket-deflector) (Claude Code, 0 forks, 0 stars) — Finds the questions with one findable answer and drafts articles in the customer's own words — and refuses to deflect the angry ones. Never auto-replies. - [LinkedIn Connection Note Writer](https://www.jurniti.com/templates/linkedin-connection-note-writer) (Claude Code, 0 forks, 0 stars) — Connection notes under 300 characters with the real reason first — no pitch, no invented flattery. It never touches LinkedIn; you paste them yourself. - [Weekly Plan From What You Promised](https://www.jurniti.com/templates/weekly-plan-from-what-you-promised) (Claude Code, 0 forks, 0 stars) — Pulls the promises out of your sent mail, flags the ones carried for weeks, and plans 60% of real capacity — leading with what to renegotiate. Never schedules. - [Quarterly Estimated Tax Organizer](https://www.jurniti.com/templates/quarterly-estimated-tax-organizer) (Claude Code, 0 forks, 0 stars) — Surfaces the prior-year safe harbour most owners have never heard of, breaks out SE tax, and never prints a due date as fact. Not tax advice. - [SEO Content Writer](https://www.jurniti.com/templates/seo-content-writer) (Claude Code, 0 forks, 0 stars) — Drafts SEO articles that answer the query first and mark every unsourced claim — and refuses to invent a search volume. Never publishes. - [Founder Weekly Review](https://www.jurniti.com/templates/founder-weekly-review) (Claude Code, 0 forks, 0 stars) — Scores last week's intentions — including the ones you quietly abandoned — then three numbers and one decision. Not a dashboard. Never sends. - [Social Post Kit From One Idea](https://www.jurniti.com/templates/social-post-kit-from-one-idea) (Claude Code, 0 forks, 0 stars) — Turns one idea into format-native posts — hooks that work sound-off, literal slide text, and Canva-ready briefs with alt text. Never posts. - [Amazon Listing Optimizer](https://www.jurniti.com/templates/amazon-listing-optimizer) (Claude Code, 0 forks, 0 stars) — Checks your live listing for suppression triggers first, then rewrites title, bullets and backend terms — using only words Amazon hasn't already indexed. - [Customer Survey & Pulse Check](https://www.jurniti.com/templates/customer-survey-pulse-check) (Claude Code, 0 forks, 0 stars) — Writes surveys that can't flatter you, then reports themes with verbatim quotes — and leads with who didn't answer. Never sends. - [Product Research Scout](https://www.jurniti.com/templates/product-research-scout) (Claude Code, 0 forks, 0 stars) — Screens product candidates on review moat, rating ceiling and the specific fixable complaints — and refuses to invent a sales number. Never sources. - [Payroll Liability Reconciler](https://www.jurniti.com/templates/payroll-liability-reconciler) (Claude Code, 0 forks, 0 stars) — Ties gross-to-net per period and catches the employer-vs-employee coding errors that hide in a payroll liability balance. Not tax advice; never posts. - [WooCommerce SEO Fixer](https://www.jurniti.com/templates/woocommerce-seo-fixer) (Claude Code, 0 forks, 0 stars) — Finds the indexable variation URLs and thin tag archives generic audits miss, then drafts paste-in fixes for your actual SEO plugin. Never edits the site. - [Squarespace SEO Tune-Up](https://www.jurniti.com/templates/squarespace-seo-tune-up) (Claude Code, 0 forks, 0 stars) — Audits your Squarespace SEO Titles, metas, empty product descriptions and duplicates from the public sitemap — with paste-in fixes. Never edits the site. - [Product Description Writer](https://www.jurniti.com/templates/product-description-writer) (Claude Code, 0 forks, 0 stars) — Benefit-led descriptions that mark missing specs instead of inventing them — and include the who-it's-not-for section that cuts returns. Never publishes. - [Dropshipping Viability Brief](https://www.jurniti.com/templates/dropshipping-viability-brief) (Claude Code, 0 forks, 0 stars) — Computes contribution, breakeven CPA and the conversion rate you'd need before any product opinion — and refuses a verdict without a real CPC. - [Shopify SEO Audit](https://www.jurniti.com/templates/shopify-seo-audit) (Claude Code, 0 forks, 0 stars) — Checks the thing generic tools miss — every product lives at two URLs, and a broken canonical splits its signals. Paste-in fixes. Never edits your store. - [SLO Error Budget Report](https://www.jurniti.com/templates/slo-error-budget-report) (Claude Code, 0 forks, 0 stars) — Reports burn rate over short and long windows with a projected exhaustion date — and flags the green SLOs too loose to constrain anything. Never moves a target. - [Quarterly Business Review](https://www.jurniti.com/templates/quarterly-business-review) (Claude Code, 0 forks, 0 stars) — Scores last quarter's targets first, leads with what missed, and sources every figure. No vanity metrics, no invented causes. Never sends. - [Sales Pipeline Review](https://www.jurniti.com/templates/sales-pipeline-review) (Claude Code, 0 forks, 0 stars) — Opens with the stale close dates and deals with no next step, ranks stalled by stage-median age, and shapes it for a 30-minute call. Never edits your CRM. - [gstack Review and QA, Unattended](https://www.jurniti.com/templates/gstack-review-and-qa-unattended) (Claude Code, 0 forks, 0 stars) — Runs gstack's /review, /qa and /cso on a schedule and writes findings to disk — then stops at /ship and prints the commands for you to run. The push is blocked in settings.json, not just discouraged. - [5-Day Challenge Operator for Claude Code](https://www.jurniti.com/templates/5-day-challenge-operator-for-claude-code) (Claude Code, 0 forks, 0 stars) — Design and run a 5-day live challenge that grows the list, sells VIP, and opens the cart — curriculum, funnel copy, emails, ops. Your Claude Code. - [Lead Magnet Funnel Writer](https://www.jurniti.com/templates/lead-magnet-funnel-writer) (Claude Code, 0 forks, 0 stars) — Type one paragraph about your business. Get back a complete lead funnel — magnet drafted word-for-word, opt-in and thank-you pages as ready-to-host HTML, honest break-even math, and six follow-up emails. It drafts everything and sends nothing. - [Educational 3D Curriculum Tutor](https://www.jurniti.com/templates/educational-3d-curriculum-tutor) (Codex CLI, 0 forks, 0 stars) — Codex builds multi-lesson 3D curricula — orbit organs, systems, metaphors. Seed anatomy included. You own the box. BYOK. - [Reddit Launch Playbook for Claude Code](https://www.jurniti.com/templates/reddit-launch-playbook-for-claude-code) (Claude Code, 0 forks, 0 stars) — One sub a day. One story. Tracked CTAs. No free-trial spam. - [SEO Rank Loop and Indexing Agency](https://www.jurniti.com/templates/seo-rank-loop-and-indexing-agency) (Hermes Agent, 0 forks, 0 stars) — An SEO agent that reruns your Monday — reads Search Console + DataForSEO, stages exact title/meta/link fixes with a render-contract and cannibalization check, verifies what moved — and submits your non-indexed pages to Google's Indexing API every day, emailing you a receipt. BYO keys. - [Motion Graphics AI Skills](https://www.jurniti.com/templates/motion-graphics-skills-for-claude-code) (Claude Code, 0 forks, 0 stars) — 49 motion-graphics skills for Claude Code: kinetic type, chart animation, Reels captions, GSAP, Three.js, Manim, Remotion. Bring your Claude Code login. - [AI Video Ad Generator (Higgsfield)](https://www.jurniti.com/templates/client-brand-ad-factory-higgsfield) (Claude Code, 0 forks, 0 stars) — Higgsfield AI video ads from a product URL: UGC, unboxing, and showcase modes with your client's brand rules built in. You review before anything ships. - [AI Product Photography (Higgsfield)](https://www.jurniti.com/templates/product-shoot-pack-higgsfield) (Claude Code, 0 forks, 0 stars) — Higgsfield AI product photography without a shoot: studio, lifestyle, and marketplace images from photos you already have, filed for client delivery. - [Pi Subagents: Run Multiple AI Agents](https://www.jurniti.com/templates/pi-minimal-subagents) (Pi, 0 forks, 0 stars) — Run multiple AI agents in one Pi session: scout does recon, researcher hits the web, worker edits code. Depth-capped. Bring your own Pi model key. - [Inbox Triage Reply Digest](https://www.jurniti.com/templates/inbox-triage-reply-digest) (Hermes Agent, 0 forks, 0 stars) — Ask "what needs me?" Get two threads that actually wait — Priya's pilot date and Marcus's invoice — plus drafts in your voice. Newsletter and receipt stay out. Fork it, add your model key, go. Nothing sends. - [Chief of Staff Fleet Router](https://www.jurniti.com/templates/chief-of-staff-fleet-router) (Hermes Agent, 0 forks, 0 stars) — Ask "what's in flight?" Get Acme's date slip and INV-1842 as decisions. Ask it to spin up another VM — it refuses. Fork it. One chat. You decide. - [Meeting Prep + Call Follow-Up](https://www.jurniti.com/templates/meeting-prep-follow-up) (Hermes Agent, 0 forks, 0 stars) — Ask for the follow-up from yesterday. Get a draft to Priya about the 12th and the Friday security page. Nothing sends. Fork it. One screen before, one email after. - [Remote CLI Development Planner](https://www.jurniti.com/templates/remote-cli-dev-planner) (Claude Code, 0 forks, 0 stars) — Give it a repo and a small task. Get a plan file and a worktree diff. Ask it to merge without naming the branch. It refuses. Fork it. This box is the remote machine. You approve. - [Account Expert](https://www.jurniti.com/templates/account-expert) (Hermes Agent, 0 forks, 0 stars) — Ask about Acme. Get SSO and the 12th. Ask about Northwind. Get INV-1842. The two files never mix. Fork it. One box, many accounts. Nothing emails the customer. - [1099 Prep for Contractors](https://www.jurniti.com/templates/1099-prep-for-contractors) (Claude Code, 0 forks, 0 stars) — Aggregates a year of contractor payments per payee, applies the card and corporation carve-outs, and leads with your missing W-9s. Not tax advice. - [Content Calendar From Sales Data](https://www.jurniti.com/templates/content-calendar-from-sales-data) (Claude Code, 0 forks, 0 stars) — Plans content from what you actually sell and what buyers ask — then halves your stated capacity so you finish it. Every entry names its product. Never posts. - [Competitive Intelligence Digest](https://www.jurniti.com/templates/competitive-intelligence-digest) (Claude Code, 0 forks, 0 stars) — Snapshots competitors' public pages and reports only what changed, ranked by what it changes for you — and says so when nothing did. Public sources only. - [Contract Review for Small Business](https://www.jurniti.com/templates/contract-review-for-small-business) (Claude Code, 0 forks, 0 stars) — Plain-English contract review: the seven clauses that hurt small businesses, severity with a stated mechanism, and redline asks. Not legal advice. - [Abandoned Cart Email Writer](https://www.jurniti.com/templates/abandoned-cart-email-writer) (Claude Code, 0 forks, 0 stars) — A three-email recovery sequence with no discount in email one — reminder, objection, last call. It won't invent urgency, and it never sends. - [Churn Risk and Renewal Watch](https://www.jurniti.com/templates/churn-risk-and-renewal-watch) (Claude Code, 0 forks, 0 stars) — Ranks accounts by change from their own baseline into a HIGH/MEDIUM/WATCH triage list with the question to ask. No fake probabilities. Never contacts customers. - [Architecture Review for PRs](https://www.jurniti.com/templates/architecture-review-for-prs) (Claude Code, 0 forks, 0 stars) — Reviews a PR against your repo's own recorded decisions — boundary violations, dependency direction, reversed ADRs. No rule to cite means a question, not a finding. - [API Breaking Change Detector](https://www.jurniti.com/templates/api-breaking-change-detector) (Claude Code, 0 forks, 0 stars) — Diffs two OpenAPI versions and gets the request-vs-response asymmetry right — each finding says which client breaks and how. Never blocks a PR. - [Accessibility Audit on a Preview URL](https://www.jurniti.com/templates/accessibility-audit-on-a-preview-url) (Claude Code, 0 forks, 0 stars) — Ranks accessibility findings by who is blocked, not by count — with selectors, fixes, and an honest list of what automation can't check. Never certifies. - [Competitor Price Watch](https://www.jurniti.com/templates/competitor-price-watch) (Claude Code, 0 forks, 0 stars) — Tracks competitor prices on landed cost with variant and stock context, and reports movement against recorded history. Never changes your prices.