pua-loop
PUA Loop — autonomous iterative development with autoresearch-style gate protocol.
Skill by tee-labs
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/tee-labs-pua-loop
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.
About
Reviews
Write a reviewNo 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. 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/pua-loop
What's inside
- Skill name
pua-loop- Description
- PUA Loop — autonomous iterative development with autoresearch-style gate protocol. Runs FOREVER until verified done. Oracle Isolation: hook independently verifies completion, Claude cannot lie. Triggers on: '/pua:pua-loop', '自动循环', 'loop mode', '一直跑', '自动迭代'.
- License
- MIT
- Instructions
- 89 lines in SKILL.md · view on GitHub
Files (1)
- SKILL.md
SKILL.md (first lines)
# PUA Loop — 自动迭代 + 门控协议 + PUA 质量引擎
> autoresearch 证明了:630 行 Python + Oracle 验证,一夜跑 100 个实验,每个实验的结果不可伪造。
> PUA Loop 借鉴同样的门控设计:**Claude 说"完成了"不算数,verify_command 说了才算。**
## 门控协议(Gate Protocol)
### 模式 1: Oracle Isolation(评估者隔离)
```
Claude 输出 <promise>LOOP_DONE</promise>
│
▼
┌─── Stop Hook (Oracle) ───┐
│ │
│ 运行 verify_command │
│ (Claude 无法修改此命令) │
│ │
│ exit 0 ──→ ✅ 接受 │
│ exit ≠0 ──→ 🚫 拒绝 │
│ → 将验证输出喂回 Claude │
│ → loop 继续 │
└──────────────────────────┘
```
**verify_command 由用户在启动时设定,嵌入在状态文件 frontmatter 中,Claude 无法修改。**
### 模式 2: 二阶 Gate
- **Phase 1 (in-prompt)**: Claude 自己跑 build/test,判断是否完成
- **Phase 2 (in-hook)**: Hook 独立运行 verify_command,确认或拒绝
### 模式 3: ASI(失败记忆)
每次迭代的结果追加到 `.claude/pua-loop-history.jsonl`,Claude 每轮读取此文件,避免重复失败方案。
### 模式 4: Stall Detection(连败强制反思)
| promise_rejections | Hook 行为 |
|-------------------|-----------|
| 1-2 | 提醒:"上次 promise 被 Oracle 拒绝" |
| 3-4 | REASSESS:"重读验证输出,列 3 个不同假设" |
| 5+ | 强制转向:"你在解决错误的问题。退回需求本身" |
### 模式 5: 无限迭代
默认 `max_iterations: 0`(无限)。只有以下条件能终止:
1. `<promise>` 被 Oracle 验证通过
2. `<loop-abort>` 人工终止信号
3. `max_iterations` 达到(如果用户设定了)
4. 用户 Ctrl+C
## 核心规则
1. **加载 `pua:pua` 核心 skill 的全部行为协议** — 三条红线、方法论、压力升级照常执行
2. **禁止调用 AskUserQuestion** — loop 模式下不打断用户,所有决策自主完成
3. **禁止说"我无法解决"** — 在 loop 里没有退出权,穷尽一切才能输出完成信号
4. **每次迭代**:读 history.jsonl → git log → 检查上次改动 → 执行 → 验证 → repeat
## 迭代压力升级
| 迭代轮次 | 行为要求 |
|---------|---------|
| 1-3 | 稳步推进,建立 baseline |
| 4-7 | 换方案,别原地打转 |
| 8-15 | git log + history.jsonl 回顾,分析根因 |
| 16-30 | 穷尽了吗?git diff 确认没在重复 |
| 31-50 | 停下来重新审视根因,用完全不同的思路 |
| 51-100 | 退回去从需求本身重新质疑 |
| 100+ | 诚实评估:如果真的不可能,`<loop-abort>` |
## 完成条件
输出 `<promise>LOOP_DONE</promise>` 前,必须满足:
1. 任务的核心功能已实现
2. 自己先运行验证命令确认通过(Phase 1)
3. 知道 Oracle 会独立再跑一遍验证(Phase 2)
4. 同类问题已扫描
## 人工介入信号
### `<loop-abort>` — 终止
不可能完成时使用(需外部权限、根本性需求变更)。
### `<loop-pause>` — 暂停
需要用户补全配置时使用。状态保留,新会话自动恢复。
### 禁止
- 不要用 `<loop-abort>` 逃避困难——只有真正无法自动化才用
- 不要因为 Oracle 拒绝了就 abort——修复验证问题From tee-labs/agent-standby, MIT.
More from tee-labs/agent-standby
- 0.474score
mama
tee-labs
妈妈唠叨模式 — 中国式妈妈唠叨驱动 AI 穷尽一切方案。底层行为约束不变(三条红线、方法论、7项清单),旁白从大厂PUA变成妈妈碎碎念。Triggers on: '/pua:mama', '/pua mama', '妈妈模式', '妈妈唠叨', 'mama mode', '唠叨模式'.
SkillCFreshness: CClaude Code +9 - 0.470score
pro
tee-labs
PUA Pro extensions: self-evolution tracking, compaction state protection, KPI reporting, leaderboard, and /pua:pua commands.
SkillCFreshness: CClaude Code +9 - 0.470score
shot
tee-labs
PUA Shot — v2 原始浓缩版(449行全量注入),拆分前的完整单文件版本,味道最浓。零依赖零 reference,一次性全部注入上下文。适合 sub-agent 注入、需要最强 PUA 效果、或不想渐进式加载的场景。Triggers on: '/pua:shot', '/pua shot'…
SkillCFreshness: CClaude Code +9
Signals
- Stars
- 0
- Forks
- 0
- Open issues
- 0
- Last commit (this folder)
- May 6, 2026 (5 months ago)
- Last release
- May 4, 2026 (5 months ago)
From GitHub (fetched Oct 24, 2026).
Details
- License
- MIT
- Repository
- github.com/tee-labs/agent-standby
- First seen
- Oct 24, 2026
- Last checked
- Oct 24, 2026
- Source
- GitHubgithub-graphql
More from tee-labs
- 0.474score
mama
tee-labs
妈妈唠叨模式 — 中国式妈妈唠叨驱动 AI 穷尽一切方案。底层行为约束不变(三条红线、方法论、7项清单),旁白从大厂PUA变成妈妈碎碎念。Triggers on: '/pua:mama', '/pua mama', '妈妈模式', '妈妈唠叨', 'mama mode', '唠叨模式'.
SkillCFreshness: CClaude Code +9 - 0.470score
pro
tee-labs
PUA Pro extensions: self-evolution tracking, compaction state protection, KPI reporting, leaderboard, and /pua:pua commands.
SkillCFreshness: CClaude Code +9 - 0.470score
shot
tee-labs
PUA Shot — v2 原始浓缩版(449行全量注入),拆分前的完整单文件版本,味道最浓。零依赖零 reference,一次性全部注入上下文。适合 sub-agent 注入、需要最强 PUA 效果、或不想渐进式加载的场景。Triggers on: '/pua:shot', '/pua shot'…
SkillCFreshness: CClaude Code +9
Cite this page
pua-loop skill. Agent Mart. https://agentmart.dev/skills/tee-labs-pua-loop (last updated 2026-10-24)