codex-autoresearch

Run repeated, measured Git experiments toward a numeric target; keep improvements and revert failures. Use for autonomous optimization or managing a…

执行命令写文件严重 0 · 高危 27leo-lilinxiao/codex-autoresearch

它会碰到什么

扫了多少34 个文本文件，246 KB

它会碰到什么执行命令写文件

命中总数28 处

命中统计严重 0 · 高 27 · 中 1 · 低 0

逐条看命中（27 条严重或高危）

高 scripts/autoresearch_core.py:616exec-spawn
```
completed = subprocess.run(
```
高 scripts/autoresearch_core.py:665exec-spawn
```
completed = subprocess.run(
```
高 scripts/autoresearch_core.py:719exec-spawn
```
completed = subprocess.run(
```

高 scripts/autoresearch_core.py:870exec-spawn

process = subprocess.Popen(command, **popen_kwargs)

高 scripts/autoresearch_core.py:1025exec-spawn
```
process: subprocess.Popen[Any],
```
高 scripts/autoresearch_core.py:1032exec-spawn
```
completed = subprocess.run(
```

高 scripts/autoresearch.py:986exec-spawn

child: subprocess.Popen[str] | None = None

高 scripts/autoresearch.py:1072exec-spawn

child = subprocess.Popen(command, **worker_kwargs)

高 scripts/autoresearch.py:1233exec-spawn
```
process: subprocess.Popen[Any],
```

高 scripts/autoresearch.py:1325exec-spawn

process = subprocess.Popen(command, **popen_kwargs)

高 scripts/run_skill_e2e.sh:96exec-spawn
```
subprocess.check_output(
```
高 scripts/run_skill_e2e.sh:164exec-spawn
```
subprocess.check_output(
```
高 scripts/run_skill_e2e.sh:173exec-spawn
```
subprocess.check_call(
```
高 scripts/run_skill_e2e.sh:203exec-spawn
```
subprocess.check_output(
```
高 scripts/run_skill_e2e.sh:231exec-spawn
```
subprocess.check_output(
```
高 scripts/run_skill_e2e.sh:251exec-spawn
```
dirty = subprocess.check_output(
```
高 scripts/run_skill_e2e.sh:258exec-spawn
```
subjects = subprocess.check_output(
```
高 tests/test_autoresearch.py:50exec-spawn
```
return subprocess.run(
```
高 tests/test_autoresearch.py:59exec-spawn
```
return subprocess.run(
```

高 tests/test_autoresearch.py:640exec-spawn

"status = json.loads(subprocess.check_output([sys.executable, script, 'status', '--repo', str(repo)], text=True))\n"

高 tests/test_autoresearch.py:644exec-spawn

"subprocess.check_call([sys.executable, script, 'finish', '--repo', str(repo), '--description', f'reduce value to {next_value}'])\n",

高 tests/test_autoresearch.py:732exec-spawn

"    subprocess.check_call(['git', 'commit', '--allow-empty', '-m', 'metric side effect'])\n"

高 tests/test_autoresearch.py:828exec-spawn
```
process = subprocess.Popen(
```
高 tests/test_autoresearch.py:862exec-spawn
```
real_popen = subprocess.Popen
```

高 tests/test_autoresearch.py:864exec-spawn

def fail_controller_only(command: object, *args: object, **kwargs: object) -> subprocess.Popen:

高 tests/test_autoresearch.py:900exec-spawn

controller = subprocess.Popen([sys.executable, "-c", "import time; time.sleep(30)"])

高 tests/test_autoresearch.py:961exec-spawn
```
child = subprocess.Popen(
```

这一栏是扫描器报的事实，不是结论。命中多不等于有毒（安全工具、规则库、示例脚本本来就会包含危险写法），命中少也不等于干净。它和你手上的凭据、文件、网络有什么关系，需要你自己看。

技能内容

Codex Autoresearch

Improve a repository through repeated, reversible experiments:

hypothesize -> change -> measure -> learn -> keep or revert -> repeat

Codex chooses hypotheses and makes code changes. The control script owns measurement, commits, rollback, and the event history.

Respond To The Request

Resolve <control> to this skill's own scripts/autoresearch.py; do not assume it is installed in the target repository.

For a status or results request, run the corresponding command directly:

| Request | Command |

|---|---|

| Status | python3 <control> status --repo <repo> |

| History | python3 <control> history --repo <repo> |

| TSV export | python3 <control> history --repo <repo> --format tsv |

| HTML report | python3 <control> report --repo <repo> |

Report the requested result without starting or resuming experiments. Return the generated path for an HTML report.

Before starting or resuming an experiment, read the applicable workflow unless it is already in context:

[Workflow](references/workflow.md): starting, continuing, resuming, or changing an experiment.
[Background](references/background.md): launching or controlling a detached run.

Experiment Boundary

One run owns one Git repository, one numeric metric, one confirmed target, and approved path scopes.
For a new foreground run, initialize successfully before creating its Goal; the Goal identifies the returned run id. If initialization reports complete, no Goal is needed.
Use finish to finalize each coherent experiment. Do not manually commit, revert, or edit run artifacts during an active run.
Only a verified target can mean complete; an iteration limit, error, or external blocker has its own status.
Validate state through the control script on entry or resume. Use recorded experiments rather than guessing from conversation.
Surface command and state errors with their diagnostic paths. Never bypass validation, fabricate metrics, or reconstruct missing events.

Foreground continuation belongs to the official Codex Goal. Background continuation belongs to the detached controller. An already launched background worker follows its supplied experiment contract, without starting a new launch flow.

想直接用这个技能？

本站把开放许可（MIT / Apache 等）的技能按仓库打包整理到网盘，点一下转存到你自己的网盘，不用一个个从 GitHub 拉。许可未声明的技能只给原始仓库链接，不打包。

去原始仓库

它属于哪个仓库

仓库leo-lilinxiao/codex-autoresearch

星标★ 2,542

本站分层T2

该仓技能数1

原文件路径SKILL.md