Skip to content
Agentmart
Submit a tool

harness-feedback

执行 F1-F5 结构化反馈处理流程,从问题中提炼规则并写入约束系统。Use when receiving bug reports, user feedback, review findings, or test failures.

Skill by duoglas

IndexedMIT★ 39
Write a review

Chalk Labs

Need help deploying harness-feedback?

Chalk Labs sets up agents like this inside real teams: connected to your tools, tested on your data, with a human in the loop where it matters.

Book a call

Install

Works with Claude Code, Claude apps, Codex, Cursor, Gemini CLI, GitHub Copilot, VS Code, OpenCode, Goose, Windsurf. Read the repository before you give any tool access to your data.

npx agentmarket install skill/duoglas-harness-feedback

Asks which of your tools to add it to (Claude Code, Codex, Cursor, VS Code, Windsurf, Gemini CLI, Claude Desktop) and shows the change first. See https://agentmart.dev/cli.

Get the weekly diff

The weekly diff: one email a week with new agent tools, notable updates and grade changes, plus security alerts for tools you follow. Double opt-in, one-click unsubscribe.

We use your email only for this newsletter. Privacy

About

执行 F1-F5 结构化反馈处理流程,从问题中提炼规则并写入约束系统。Use when receiving bug reports, user feedback, review findings, or test failures.

Use cases: testing

No reviews yet. Installed it? Reviews come from verified installers only. How reviews work

Permissions

What this skill can do and what it asks for, from its SKILL.md and code. Checked Oct 25, 2026.

Tools
No tools found
Secrets
None asked for
Install scripts
None found

Network hosts it talks to 1

  • codeload.github.com · found in install command

Files it touches

  • ~/.claude/skills/harness-feedback

What's inside

Skill name
harness-feedback
Description
执行 F1-F5 结构化反馈处理流程,从问题中提炼规则并写入约束系统。Use when receiving bug reports, user feedback, review findings, or test failures.
License
MIT
Instructions
123 lines in SKILL.md · view on GitHub

Files (1)

  • SKILL.md

SKILL.md (first lines)

# Harness Feedback

结构化反馈处理——从问题到规则的闭环。

## 何时使用

- 收到用户反馈("这个不对")
- Review 发现问题
- 测试失败
- 任何"需要修"的情况
- 用户说"处理反馈"或"F1-F5"

## 执行流程

### Step 0: 收集反馈信息

如果用户在 `/harness-feedback` 后面直接附了内容,用那个内容。否则依次询问:

1. **问题是什么?** — 等待用户描述
2. **期望行为是什么?** — 等待用户回答

收集到信息后,进入 F1-F5。

### F1: 记录原话

原样记录问题描述,不解读、不简化、不猜测意图。

```markdown
## 反馈记录
- **来源**: [用户反馈 / Review / 测试失败 / 自查]
- **原话**: "[原样记录]"
- **时间**: [YYYY-MM-DD]
- **严重性**: [阻断 / 重要 / 建议]
```

### F2: 分类问题层级

判断:这个问题只出现在一个地方,还是可能出现在多处?

| 层级 | 判断依据 | 修改位置 |
|------|---------|---------|
| 规则层 | 多处可能出现同样问题 | docs/constraints.md |
| 工具层 | 工具逻辑 bug | src/(通过 Agent) |
| 配置层 | 配置不当 | 配置文件 |
| 实例层 | 仅此一处 | 具体文件 |

### F3: 提炼规则(关键步骤,禁止跳过)

从具体实例抽象为通用规则:

```
❌ "把登录按钮的 padding 改成 16px"
✅ "所有可交互元素在移动端最小触控区域为 44x44pt,padding ≥ 12px"
```

提炼方法:
1. 问"还有哪些地方可能有同样问题?"
2. 具体值 → 通用规则("16px" → "≥ 12px")
3. 具体对象 → 类别("登录按钮" → "所有可交互元素")
4. 同类问题归一条规则

### F4: 写入文件

规则层问题写入 `docs/constraints.md`:

```markdown
## C-{AREA}-{NN}: {规则标题}

{规则描述}

- **WHY**: {为什么需要这条规则}
- **违反后果**: {不遵守会怎样}
- **来源**: VH-{NN}({日期} {问题描述})
```

同时在 Violation History 中记录:

```markdown
| VH-{NN} | {日期} | {发生了什么} | {根因} | C-{AREA}-{NN} |
```

**先写规则,再派 Agent。**

### F5: 派 Agent 修复

```markdown
修复 C-{AREA}-{NN} 违规。
读取 docs/constraints.md 中该约束的完整描述。
扫描 {范围} 中所有不符合该约束的地方,全部修复。
修复后运行 {测试命令} 确认通过。
```

Agent prompt 必须引用 Constraint ID。

## 输出

每次 F1-F5 完成后,输出处理摘要:

```
反馈处理完成
============
问题: {一句话描述}
层级: {规则层/工具层/配置层/实例层}
规则: C-{AREA}-{NN} — {规则标题}
违规记录: VH-{NN}
修复: Agent 已按规则修复并自验通过
```

## AI 工具内测试准出协议

只要任务涉及代码变更,AI 不能等用户提醒才验证。按下面顺序做:

1. 先判断风险等级:low / medium / high / release。
E2E PASS 不等于充分;如果只是 echo ok、空脚本、只 smoke、或没覆盖本次风险,用户报告要先说“现在还不能交付”,再说明测到了什么、没测到什么、下一步补什么;机器状态放最后,例如:机器状态:NOT_SUFFICIENT。DEGRADED 不能说成 PASS。
2. 识别测试能力:单测、lint、coverage、E2E、runtime smoke。
3. medium / high / release 任务必须有 E2E 证据;只有 low 小改可以不强制 E2E;找不到 E2E 入口时,先生成计划或只问一个具体启动问题。
4. VERIFY 阶段必须产出 fresh evidence;没有 READY evidence 不能说“完成了”。
5. 测试失败时进入修复 loop:一轮只修一个失败点,重跑最小测试,最多 3 轮;没进展就停下来说明卡点。
6. 报告必须说人话:先说现在能不能交付,再说测到了什么、没测到什么、下一步补什么;不能只贴日志,也不要用 READY/NOT_READY/NOT_SUFFICIENT 开头。机器状态如果必须出现,放最后。不能把 DEGRADED 说成 PASS。

From duoglas/simple-harness-kit, MIT.

More from duoglas/simple-harness-kit

Signals

Stars
39
Forks
3
Open issues
0
Last commit (this folder)
Jun 5, 2026 (5 months ago)
Last release
Oct 9, 2026 (25 days ago)

From GitHub (fetched Oct 24, 2026).

Details

License
MIT
First seen
Oct 24, 2026
Last checked
Oct 24, 2026

More from duoglas

Done for you

Rather not set it up yourself?

Chalk Labs scopes, builds and runs agents for your team: the task audit, the eval harness, the integrations and the weekly ops report. Fixed scope, priced before we start.

See how we build AI agents →Read the case studies

Chalk Labs is a two-founder AI and growth studio. We build AI agents, CRMs and MVPs for Web3 and AI companies, and have shipped 50+ product and token launches.

launches shipped
50+
launches shipped
media placements
500+
media placements
founders on every account
2
founders on every account

Press we've placed: Forbes · CoinDesk · Decrypt · Cointelegraph · The Block

Rahil Jain, Co-founder, Chalk Labs. Ex-Goldman Sachs engineer. You talk to the person who builds it.

Get harness-feedback running in your business

Tell us what you want to automate. A founder replies within one working day with a plan, not a pitch deck.

Goes to the Chalk Labs founders, not the newsletter. Privacy

Need help deploying harness-feedback?

CallGet help
Cite this page

harness-feedback skill. Agent Mart. https://agentmart.dev/skills/duoglas-harness-feedback (last updated 2026-10-24)