agentleFS
Sign inStart free
loganware05loganware05/captains-compass-cursor/.cursor/skills/autonomy-budget/SKILL.md

autonomy-budget

Skill2 starsChanged 33 days agoMITView on GitHub

Tracks per-plan autonomy budgets and stops with a Budget Stop Report when limits are hit

  • Reads credentials

1 other project uses a version of this.

Adopted by

Most recently taken up by

The file

We split this skill into the rules it gives an agent

9 rules. About 0 are common, 5 are distinctive, and 4 are very specific to this project.

  • ≈0

    Common · estimate

    Many files like this have a rule of this kind.

  • ≈5

    Distinctive · estimate

    This author's own choices. Worth a look.

  • 4

    Very specific

    Tied to their project's names, files or tools. Probably skip.

  • Common
  • Distinctive
  • Very specific
  1. Very specific:

    When the plan status becomes APPROVED, create .agent/budgets/<plan-id>.md from templates/agent/BUDGET_LEDGER.md (or copy from …

    line 21
  2. Distinctive:

    After each implementation + validation cycle, update the ledger:

    line 25
  3. Very specific:

    If validation failed, increment failed_validation_cycles

    line 27
  4. Very specific:

    Update elapsed_minutes and estimated_cost_usd (label estimates clearly)

    line 28
  5. Distinctive:

    Before starting another cycle, compare used vs maximum for every limit.

    line 30
  6. Distinctive:

    If any limit is reached or exceeded:

    line 31
  7. Very specific:

    Write a Budget Stop Report under .agent/evidence/<plan-id-or-slug>/BUDGET_STOP_REPORT.md using templates/agent/BUDGET_STOP_REPORT.md

    line 33
  8. Distinctive:

    Ask the Captain for the next action (raise budget, narrow scope, or abandon)

    line 37
  9. Distinctive:

    Never raise budget limits unilaterally.

    line 41

≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.

Keep what fits and drop the rest. Your version, ready for your agents.
Files
SKILL.md

Autonomy Budget

Use this Skill when

An implementation plan is APPROVED and autonomous execution begins, or whenever an iteration / validation cycle completes, or when any budget limit may have been reached.

Inputs

  • Approved IMPLEMENTATION_PLAN.md (Autonomy Budget section)
  • Plan ID (from plan metadata)
  • Current counters (iterations, failed validation cycles, elapsed time, estimated cost)

Procedure

  1. When the plan status becomes APPROVED, create .agent/budgets/<plan-id>.md from templates/agent/BUDGET_LEDGER.md (or copy from .agent/budgets/_templates/BUDGET_LEDGER.md if installed). Fill limits from the plan’s Autonomy Budget section.
  2. After each implementation + validation cycle, update the ledger:
    • Increment iterations_used
    • If validation failed, increment failed_validation_cycles
    • Update elapsed_minutes and estimated_cost_usd (label estimates clearly)
    • Append a one-line cycle note under the log section
  3. Before starting another cycle, compare used vs maximum for every limit.
  4. If any limit is reached or exceeded:
    • Stop further implementation
    • Write a Budget Stop Report under .agent/evidence/<plan-id-or-slug>/BUDGET_STOP_REPORT.md using templates/agent/BUDGET_STOP_REPORT.md
    • Set plan/progress notes to reflect the stop
    • Ask the Captain for the next action (raise budget, narrow scope, or abandon)
  5. At phase boundaries, append a short note under .agent/sessions/ using templates/agent/SESSION_NOTE.md (or .agent/budgets/_templates/ siblings when installed). Keep machine traces under .agent/runs/ if used.
  6. Never raise budget limits unilaterally. Never weaken tests to stay under budget.

Ledger path

.agent/budgets/<plan-id>.md

Private cost notes (optional): .agent/budgets/private/ (gitignored pattern).

Output

  • Updated budget ledger
  • Budget Stop Report when stopped
  • Clear Captain-facing recommendation

Prohibited actions

  • Continuing past any budget limit without a Budget Stop Report and Captain decision
  • Hiding failed validation cycles
  • Storing secrets or .env contents in ledgers or stop reports

Download now (ZIP)

Does it work?

Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.

No reviews yet

People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.

  • It worked
  • Partly
  • It didn't
What did you use it for? e.g. Reviewing pull requests in a Rails app

Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.

Your trail

Files you open will collect here, so you can find them again.

Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.

Now do this for your own.

Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

Check your setup, free