Claude Code logo
Claude Code template
docs.claude.com

Weekly Plan From What You Promised

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.

NewNew8.9 KB snapshotStarter VM
Claude Code logo

What's inside

Harness

Claude Code

Plan

Starter

vCPU

1

Memory

2 GiB

Snapshot

8.9 KB

How it works · ~3 minutes

  1. 01 · Fork

    New isolated microVM on your subdomain — creator state included.

  2. 02 · Your keys

    Log into Claude Code with your own model credentials (BYOK).

  3. 03 · Ask it to work

    Open the terminal and give it a real job. You keep what ships.

30-day money-back on your first purchase · no free trial · keys never leave the VM

About this template

You told a customer you'd send it "by Friday". That was three weeks ago.

It's sitting in a sent email nobody re-read. Meanwhile the weekly planner you filled on Sunday was abandoned by Tuesday, because it was built from what you wanted to do at full capacity.

Commitments, not tasks

A task is what you'd like to do. A commitment is something you told another human you would do.

Breaking a commitment costs trust. Missing a task costs time. So commitments come first — extracted from your sent mail, meeting notes and threads by scanning for promise language: "I'll send", "by Friday", "let me get you".

Each becomes: who you promised, what, and by when. No date? [NO DATE PROMISED] — which is itself the finding, because the undated promise is the one that rots.

Carried weeks are the signal

It reads last week's plan and flags anything still open as CARRIED, with the count.

Send the revised quote — Dana — carried 3 weeks.

A promise carried three weeks isn't a scheduling problem. No task list ever tells you that.

Capacity is 60% of what you said you had

Applied to real hours after meetings, and stated deliberately. The other 40% goes to what always happens and never gets planned: interruptions, the thing that breaks, the reply that takes an hour.

A plan at 100% capacity fails on Tuesday and takes the whole system's credibility with it.

Over-commitment is the output, not a footnote

Most planners hide it by quietly shrinking estimates. When your dated promises alone exceed capacity, this one stops filling and leads with the real answer:

Which promise to renegotiate, with whom, today.

Telling someone early is cheap. Missing silently isn't.

Then exactly one important-not-urgent thing. Two is zero.

What you supply

Your own Claude Code login, plus sent email, notes or a brain-dump in ~/weekplan/inbox/ — and honest hours available after meetings.

No calendar or mail connection, which is also why nothing can be scheduled or sent.

Where this stops

It drafts a plan. It does not create calendar events, send email, or accept meetings.

It won't estimate how long something takes — it uses yours or writes [ESTIMATE NEEDED], because an invented estimate is how the 60% stops protecting you. It won't call a commitment unimportant, and it won't comment on a colleague's reliability.

Verified on build: the skill, the working tree, the NEVER_SCHEDULE_OR_SEND guard, the commitments-not-tasks rule, the four buckets and the over-commitment rule are present on a fresh fork. Running it on your own week is yours to do.

Inside this fork

Forking copies this template into a brand-new, fully isolated microVM on your own subdomain. Here's exactly what lands in it.

  • Claude Code agent

    The upstream harness, pre-installed — same version the creator ran.

  • Starter VM

    1 vCPU · 2 GiB RAM · 10 GiB disk.

  • Creator's /persist data

    The captured persist volume is copied byte-for-byte into your fork.

  • BYOK — your keys, your VM

    Add your model API keys after forking; they live only inside your microVM.

What this agent can do

1 skill

  • weekly-commitments

    Build a weekly plan from the promises you made to people — extracted from sent mail and notes — sized to 60% of real capacity, surfacing…

What you'll configure after forking

Secrets are scrubbed from shared templates — these are the names you supply in your agent's terminal once it boots.

Environment variables

  • ANTHROPIC_API_KEY

Your turn

Your own Weekly Plan From What You Promised, live in about 3 minutes.

Forking copies this Claude Codeagent into a brand-new, fully isolated microVM on your own subdomain — the creator's /persist state and all. Add your own keys after it boots; they never leave the box. Don't love it? Your first jurniti purchase comes with 30 days to get every cent back.

New paid VM · BYOK · 30-day money-back on your first purchase · ~3 min to provision

Starter · fork

$25/ mo

Needs 1 of your own API key