Install
$ agentstack add skill-hyx-lhj-mini-harness-using-harness ✓ scanned · ✓ verified — works with Claude Code, Cursor, and more.
Security review
✓ PassedNo issues found. Passed automated security review. · v0.1.0 How review works →
- ✓ Prompt-injection patterns
- ✓ Secret / credential exfiltration
- ✓ Dangerous shell & filesystem operations
- ✓ Untrusted network calls
- ✓ Known-malicious package signatures
What it can access
- ✓ Network access No
- ✓ Filesystem access No
- ✓ Shell / process execution No
- ✓ Environment & secrets No
- ✓ Dynamic code execution No
From automated source analysis of v0.1.0. “Used” means the capability is present in the source — more access means more to trust, not that it’s unsafe.
About
🧭 使用 Harness(工作流入口)
mini-harness 是一组 Agent Skills,本文件是工作流入口——告诉你每回合先读什么、何时调用哪个专项 Skill。细则见 [references/workflow.md](references/workflow.md)。
| 层级 | 内容 | 谁维护 | |------|------|--------| | 🏗️ 平台层 | harness/skills/、模板 | 插件 update 同步 | | 🎯 项目层 | profile/PROJECT.md、DECISIONS.md | 项目团队,update 不覆盖 |
开场(每回合必做)
动手改任何文件之前:
- 确认路径(见下表)
- 读状态 — 并行打开
PROGRESS.md、todo.md、profile/PROJECT.md;Plan / 架构 / 跨模块时加读DECISIONS.md相关主题 - 跟流程 — [workflow.md](references/workflow.md):有变更先 todo;运行时代码须 AC 已确认 后再 TDD / 实现
> ⚠️ 跳过本 Skill 或用 ~/.agents/skills/ 同名 Skill 替代 harness/skills/,容易丢状态、丢门禁。
路径
| | 仅插件(未 install) | 仓库已 install | |---|------------------------|------------------| | 本 Skill | skills/using-harness/SKILL.md | harness/skills/using-harness/SKILL.md | | 其它 Skill | skills//SKILL.md | harness/skills//SKILL.md | | 项目画像 | — | harness/profile/PROJECT.md | | 重大决策 | — | harness/DECISIONS.md | | 工具链 | — | pyproject.toml;可选 .mini-harness.json → commands.gate | | 状态 | — | PROGRESS.md、todo.md |
install / update / doctor → [lifecycle.md](references/lifecycle.md)
硬约束
这些规则存在,是因为新会话没有聊天记忆——文件才是真相源。
- 有文件变更 → 先登记
todo.md(纯只读诊断除外) - 改运行时代码 → todo 含 AC,且
tests/有对应测试 - AC 已确认 → 用户核对意图并勾选前,不写实现、不跑 TDD
- 遵守项目画像 →
PROJECT.md优先于通用默认;门禁以pyproject.toml/commands.gate为准 - 尊重重大决策 → 触及
DECISIONS.md已记录结论时,须新 todo + 用户确认,勿静默推翻 - 范围 → 不覆盖用户无关改动
- Subagent → 测试、验收、审查、提交前精炼走 subagent;主 Agent 不代劳
- 进化须确认 → 重大取舍 →
DECISIONS对应主题;可执行规则 →PROJECT.md;evolution.jsonl只追加 - 并行 → 无依赖冲突时同消息启动可并行 subagent(见 workflow 并行规则)
流程一览
📖 读状态 → [📝 Plan?] → todo + AC 核对 → 🧪 subagent(测试) → ⚡ 实现 → 🚦 门禁
→ ✅ 验收 ∥ 🔍 审查 → [🌱 进化?] → 📁 归档 → PROGRESS
提交:✨ 精炼 → 🔍 审查 → git
逐步说明 → [workflow.md](references/workflow.md)
参考
| 文档 | 内容 | |------|------| | [workflow.md](references/workflow.md) | 逐步流程、并行、Subagent、进化 | | [lifecycle.md](references/lifecycle.md) | install / update / doctor | | [host-support.md](references/host-support.md) | Cursor · Codex · Claude Code | | [skill-index.md](references/skill-index.md) | Skill 速查与误用 | | [portability.md](references/portability.md) | 可移植性边界 |
Source & license
This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.
- Author: HYX-LHJ
- Source: HYX-LHJ/mini-harness
- License: MIT
Install and usage instructions live in the source repository linked above.
Reviews
No reviews yet — be the first.
Write a review
Versions
- v0.1.0 Imported from the upstream source.