dsh-harness-audit
v0.1.0 · 1 天前发布
Audit an agent harness against the harness-evaluation criteria, with machine-enforced evidence validation.
README
dsh-harness-audit
Audits an agent harness for the failure modes that break agent loops — and
refuses any finding whose evidence isn't actually in the code.
Fifteen checks, one throwaway subagent each, and a summary written by code
rather than by a model. The plugin is self-contained: installing it installs
the criteria too.
What it actually catches
The criteria are failure modes, not style. A sample:
| | |
|---|---|
| C1 | A tool call gets no result, so the next request carries a call with no answer and the model service refuses it. |
| C9 | An operation times out after its external effect landed, so the retry does it twice — two messages sent, two charges. |
| C10 | The user cancels; the subprocess keeps running and the request is still in flight. |
| C12 | The
版本兼容矩阵
| DSH 版本 | 状态 | 说明 |
|---|---|---|
dsh@0.1.0-rc.6 | ✅ 兼容 | cordis patch ✅ |
dsh@0.1.0 | ✅ 兼容 | 标准 patch ✅ |
dsh@0.2.0 | ⚠️ 待验证 | 待实测(API 可能在变) |
dsh@0.3.0+ | ❓ 未知 | 尚未发布 |
💡 兼容矩阵每周末自动跑实测,欢迎 PR 修正
元数据
- License
- MIT
- npm
- dsh-harness-audit
- Repo
- git+https://github.com/Leeaoyin/dsh-agent-harness-audit.git
- Engines
{}- Weekly DL
- 0
- Monthly DL
- 0
- dsh.bundle
{"bundle":{"patch":"./cordis.patch.yml"}}