/g The Contract Codex ← 返回动态 Catalog

Refactor Until Stop Conditions Met

Let the agent iterate on a refactor continually until one of several predefined stop conditions is met, resolving blockers so each run works longer.

难度:intermediate 分类:refactor 来源:How to achieve greater code refactoring automation? - OpenAI Community

完整 Prompt(可直接复制)

/goal
GOAL:
Complete Refactor Until Stop Conditions Met for a refactoring task: Let the agent iterate on a refactor continually until one of several predefined stop conditions is met, resolving blockers so each run works longer.

CONTEXT:
- Before editing, read the nearest AGENTS.md/CLAUDE.md, current issue or PLAN.md, and any failing logs already in the repo.
- Inspect the target module, call sites, public API, tests, and compatibility notes.
- Establish a baseline by running or locating evidence for: `pytest && git status --short`.

CONSTRAINTS:
- Keep the scope limited to this goal; do not expand into unrelated cleanup.
- Do not weaken tests, delete assertions, or mask errors to make verification pass.
- Respect the repository's AGENTS.md/CLAUDE.md instructions and existing patterns.
- Do not change public APIs, data formats, or user-visible behavior unless the goal requires it.
- Keep behavior characterization tests before large internal changes.

DONE WHEN:
- The implementation or documentation directly satisfies: Let the agent iterate on a refactor continually until one of several predefined stop conditions is met, resolving blockers so each run works longer.
- The verification command or evidence path succeeds: `pytest && git status --short`.
- The final diff is scoped to the relevant files and has no unrelated formatting churn.

VERIFY:
- Run `pytest && git status --short` or the closest repo-local equivalent if the exact command is not available.
- Capture before/after evidence for the behavior, metric, report, or artifact involved.
- If verification cannot run locally, stop and report the missing dependency instead of guessing success.

OUTPUT:
- Summarize changed files, key decisions, verification output, and remaining risks.
- Include any follow-up that is required for production rollout or human review.

STOP RULES:
- Pause if secrets, production access, stakeholder decisions, or destructive data operations are required.
- Pause after three failed fix attempts on the same symptom and challenge the root-cause hypothesis.
- Do not mark the goal complete until the current repository state has been audited against DONE WHEN.

来源与证据

原始来源: How to achieve greater code refactoring automation? - OpenAI Community

证据摘要: I suggested we allow Codex to iterate continually until one of numerous stop conditions was met.; source: How to achieve greater code refactoring automation? - OpenAI Community; type: public-forum; verification: pytest && git status --short