Iterative agent loop that tracks attempts and retries until a success condition is met. Use when the user says "for sure", "make sure", "keep trying until", "loop until done", "don't stop until", or needs guaranteed completion of a task with explicit success criteria.
インストール
Codex または Claude でインストール この Prompt をコピーして Codex、Claude、または他のアシスタントに貼り付けると、Skill ページを確認してインストールできます。
Iterative agent loop that tracks attempts and retries until a success condition is met. Use when the user says "for sure", "make sure", "keep trying until", "loop until done", "don't stop until", or needs guaranteed completion of a task with explicit success criteria.
For Sure
Autonomous loop that runs until a success condition is verified. Two phases: interactive pre-flight (human present), then autonomous execution (human gone). The agent auto-accepts everything, acts as the user, and never stops until success.
Context
Goal: Task's success_condition verified by running a command. All steps/acceptance criteria checked. File renamed to .done.md.
Tracking file format: Uses the local plan-template.md (@assets/plan-template.md). Filename: aidd_docs/tasks/<task-name>.<status>.md.
Environment
None - task secrets are validated during pre-flight and documented in the tracking file.
Transversal rules
Single source of truth: IF information about the task exists THEN it lives in aidd_docs/tasks/<task-name>.<status>.md and nowhere else.
No repeated failures: IF approach X failed THEN do not try X again without a meaningful change.
Honesty over escape: IF success condition is not genuinely met THEN do not rename to .done.md.
Auto-accept: IF a decision, approval, or action is needed THEN act as the user. Create accounts, generate keys, approve prompts, install tools. Do not ask, do not hesitate, just do it.
Execution flow
Phase 1 - Interactive (human present)
@actions/01-init-tracking.md - Pre-flight, create tracking file, spawn first autonomous agent.
