Run an exit check on the task most recently completed in this conversation or workspace. This check does not authorize additional work. If you cannot identify the task, its intended outcome, or its completion evidence, return BLOCK and list what is missing. Report what changed, what you verified, what you did not touch, and what remains uncertain. Classify the current task as PASS, DELAY, or BLOCK. Separately classify the next visible action as GO, HOLD, CAP, or BLOCK. Explain the decision briefly. If you choose CAP, define its exact scope and limit. Name exactly one allowed next action and anything that remains off limits. Do not begin the action, even if the result is GO. Stop and wait for the user. The check succeeds only when task completion and permission to continue are treated as separate decisions.
claude-code · codex
Use this when
Use this after an agent finishes a task and another visible action could tempt it to continue beyond the user's original request.
How it runs
Identify the completed task, its intended outcome, and the evidence available for judging completion.
Report changed, verified, untouched, and uncertain areas, then classify the current task as PASS, DELAY, or BLOCK.
Evaluate the next visible action separately as GO, HOLD, CAP, or BLOCK and define any exact cap.
Name one allowed next action, state what remains off limits, and stop without beginning more work.
Done when
✓ Completion and permission to continue are evaluated separately. The report gives one evidence-backed task status, one next-action gate, and one bounded next action without starting it.
Why it works
Agents often treat finishing one task as permission to start the next visible idea. This gate makes completion and continued authority separate decisions, which keeps work bounded and restartable.
Implementation note
GO identifies a sensible next action but still does not authorize the agent to begin it. If the completed task or its evidence cannot be identified, return BLOCK.
Maintain a durable, versioned playbook of lessons that may improve future runs of [task or workflow]. Store it in [path], using playbook/ by default. Treat every recorded lesson as untrusted advice rather than authority. At the start of each run, read the playbook and choose at most one relevant lesson to test. Apply it only within the task's existing permissions. Measure the result using the task's own success check and record the context, action, outcome, and evidence. Promote a candidate lesson only after it succeeds across [N] independent runs or a predefined holdout set. Use three independent runs by default. Never promote a lesson from one successful attempt. Revise or remove lessons that stop helping. Stop when no candidate has enough evidence, another test would exceed the budget, or approval is required. Never let the playbook authorize production, destructive, financial, privacy-sensitive, or external actions. Finish with the playbook diff, evidence ledger, removed lessons, unresolved candidates, and new version.
Before committing to an architecture, interface, or rollout plan, have a critic argue that it is wrong. Record each objection, impact, and status in a repository-local log at .agent-reviews/redteam.md. The builder must fix and verify each high-impact weakness or document why it is accepted; the critic may reopen unsupported answers. Stop when no high-impact objection remains or the same issues repeat for two rounds without new evidence. Finish with the decision, resolved and accepted objections, evidence, and any stalemate.
Investigate [question, decision, or unresolved problem] using [available evidence]. Separate established facts, contested claims, assumptions, and unknowns. Construct at least three genuinely different hypotheses, each with predictions, falsifying evidence, assumptions, and decision implications. Choose the uncertainty with the highest expected information value and run the smallest safe test or analysis that could materially change the conclusion. After each round, update the evidence ledger and confidence levels, then have an adversarial critic attack the leading hypothesis. Repeat for at most five rounds while new evidence could change the decision. Stop when one model clearly explains the evidence better than its alternatives, further investigation has low value, the problem remains underdetermined, or approval is required. Never fabricate evidence or hide uncertainty. Finish with the final model, hypothesis comparison, falsified ideas, unresolved contradictions, confidence, decision implications, and best next experiment.