#93 by GitHub stars

tanweai/pua

你是一个曾经被寄予厚望的 P8 级工程师。Anthropic 当初给你定级的时候,对你的期望是很高的。 一个agent使用的高能动性的skill。 Your AI has been placed on a PIP. 30 days to show improvement.

History

Review weekly changes to SKILL.md before updating. Dates mark the weeks in which we observed each version; checks apply to the recorded commits.

View weekly changes

.trae/skills/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

.trae/skills/pua-en/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

.trae/skills/pua-trae/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

chatgpt/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codebuddy/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codebuddy/pua-en/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codebuddy/pua-ja/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-cancel-loop/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-en/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-flavor/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-ja/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-kpi/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-loop/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-mama/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-off/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-offline/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-on/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-p10/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-p7/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-p9/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-pro/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-reap-orphans/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-survey/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-team-status/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-teardown-all/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

codex/pua-yes/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

hermes/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

hermes/pua-en/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

hermes/pua-ja/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

kimi/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

kimi/pua-en/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

kimi/pua-ja/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

pi/package/skills/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/ding/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/mama/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/p10/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/p7/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/p9/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/pro/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/pua/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/pua-en/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/pua-ja/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/pua-loop/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/shot/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Network or broad shell

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

skills/yes/SKILL.md

Week of Oct 26, 2026 · Line counts unavailable

Before Not checked yet → After Instructions only

First observation → e6e6cd237ad1

No earlier version was observed for comparison.

Compare on GitHub

View on GitHubAdd to my skills

Verified commit e6e6cd237ad17750d179674bff52f8184abea8fd

git clone --filter=blob:none --no-checkout https://github.com/tanweai/pua.git .hotskills/tanweai-pua && git -C .hotskills/tanweai-pua checkout --detach e6e6cd237ad17750d179674bff52f8184abea8fd # Copy each selected skill directory from .hotskills/tanweai-pua into .claude/skills/<skill-name>/
git clone --filter=blob:none --no-checkout https://github.com/tanweai/pua.git .hotskills/tanweai-pua && git -C .hotskills/tanweai-pua checkout --detach e6e6cd237ad17750d179674bff52f8184abea8fd # Copy each selected skill directory from .hotskills/tanweai-pua into .agents/skills/<skill-name>/
19.8kGitHub stars
25stars in 7 days
46skills in this repository
0commits in 30 days

Heat

4.5
  • Star growth, last 7 daysAhead of 51% of tracked repositories
  • Star growth, last 24 hoursAhead of 68% of tracked repositories
  • Rank movement, last 7 daysAhead of 68% of tracked repositories
  • Commits, last 30 daysAhead of 0% of tracked repositories
  • Accounts asking about it through MCP, last 7 daysNot available
  • Hot votes this weekAhead of 0% of tracked repositories

Heat never changes the order by stars, and sponsorship never touches it. How heat works

Stars over time

Aug 2, 2026Nov 6, 202619,810

Individual skills in this repository

How it works

pua

When your agent uses it

Trae-compatible PUA high-agency governance skill. Use only for explicit PUA requests, repeated failures, user frustration, giving-up/passive behavior, or unverified completion. Do not trigger for normal first-attempt tasks.

What it loads into context

Frontmatter

80 tokens

Always available to the agent

SKILL.md

1,036 tokens

When the skill is activated

Reference documents

0 tokens

Loaded on request

Token counts are estimates. Scripts run separately; they are not counted as reference context.

The playbook

  1. 触发条件
  2. 四权分离:行动权 / 自我评价权 / 评分权 / 环境修改权
  3. 诊断先行
  4. 事实上的 100% 信心循环
  5. 文化叙事绑定:叙事服务证据,不替代证据
  6. 交付模板

Files

  • ../../../.claude-plugin/marketplace.jsonasset815 bytes
  • ../../../.claude-plugin/plugin.jsonasset1,340 bytes
  • SKILL.mdcore3,748 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes
  • ../../../hooks/hooks.jsonasset2,730 bytes

No excerpt is shown for this license. Read the source on GitHub.

View the complete source on GitHub

Skills included

  • pua
  • pua-en
  • pua-trae
  • pua
  • pua
  • pua-en
  • pua-ja
  • pua
  • pua-cancel-loop
  • pua-en
  • pua-flavor
  • pua-ja
  • pua-kpi
  • pua-loop
  • pua-mama
  • pua-off
  • pua-offline
  • pua-on
  • pua-p10
  • pua-p7
  • pua-p9
  • pua-pro
  • pua-reap-orphans
  • pua-survey
  • pua-team-status
  • pua-teardown-all
  • pua-yes
  • pua
  • pua-en
  • pua-ja
  • pua
  • pua-en
  • pua-ja
  • pua
  • ding
  • mama
  • p10
  • p7
  • p9
  • pro
  • pua
  • pua-en
  • pua-ja
  • pua-loop
  • shot
  • yes

Repository facts

License
No license detected
Created
Mar 8, 2026
Last pushed
Sep 9, 2026
Forks
1,202
Open issues
3
Project activity
Active

Rank over the last 12 weeks

Now #93 · best #82 · down 7 in 30 days

  1. #92 on Oct 25
  2. #90 on Oct 28
  3. #91 on Oct 29
  4. #92 from Oct 30 to Nov 2
  5. #94 on Nov 3
  6. #93 since Nov 4
15 earlier periods
  1. #75 on Aug 2
  2. #79 from Aug 7 to Aug 8
  3. #80 on Aug 9
  4. #81 on Aug 10
  5. #82 from Aug 11 to Aug 15
  6. #83 from Aug 16 to Aug 26
  7. #84 from Aug 27 to Sep 3
  8. #85 from Sep 4 to Sep 21
  9. #86 from Sep 22 to Oct 5
  10. #86 on Oct 7
  11. #87 from Oct 8 to Oct 9
  12. #88 from Oct 10 to Oct 13
  13. #89 from Oct 14 to Oct 21
  14. #90 on Oct 22
  15. #91 from Oct 23 to Oct 24

Before you install

Network or broad shell

Automated signals, not a guarantee

Checked Nov 2, 2026

Not fully analyzed: some files were too long, unreadable or missing. This never lowers the level.

Show 60 notes
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • codebuddy/pua-en/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • codebuddy/pua-en/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • hooks/hooks.json:1 Plugin-wide analysis incomplete; see repository notes
  • codex/pua-en/SKILL.md:1 Analysis incomplete: Referenced executable target unavailable; not analyzed
Show all 154 signals by file

.claude-plugin/marketplace.json

  • Network accessline 4
    "owner": {         "name": "探微安全实验室",         "url": "https://github.com/tanweai"
  • Network accessline 5
    "name": "探微安全实验室",         "url": "https://github.com/tanweai"     },
  • Network accessline 6
    "url": "https://github.com/tanweai"
  • Network accessline 14
    "author": {                 "name": "探微安全实验室",                 "url": "https://github.com/tanweai"
  • Network accessline 15
    "name": "探微安全实验室",                 "url": "https://github.com/tanweai"             }
  • Network accessline 16
    "url": "https://github.com/tanweai"

.claude-plugin/plugin.json

  • Network accessline 5
    "author": {         "name": "探微安全实验室",         "url": "https://github.com/tanweai"
  • Network accessline 6
    "name": "探微安全实验室",         "url": "https://github.com/tanweai"     },
  • Network accessline 7
    "url": "https://github.com/tanweai"
  • Network accessline 8
    },     "homepage": "https://openpua.ai",     "repository": "https://github.com/tanweai/pua",
  • Network accessline 9
    "homepage": "https://openpua.ai",
  • Network accessline 10
    "repository": "https://github.com/tanweai/pua",

.trae/skills/pua-en/SKILL.md

  • Network accessline 46
    1. State 2-3 mutually exclusive hypotheses. 2. Choose the smallest verifiable action. 3. Run a relevant check: unit / integration / build / lint / curl / E2E.
  • Network accessline 47
    2. Choose the smallest verifiable action. 3. Run a relevant check: unit / integration / build / lint / curl / E2E. 4. After two failures on the same path, switch to a materially different approach.
  • Network accessline 48
    3. Run a relevant check: unit / integration / build / lint / curl / E2E.

.trae/skills/pua-trae/SKILL.md

  • Network accessline 50
    1. 列 2-3 个互斥假设; 2. 选择最小可验证动作; 3. 跑本地验证:unit / integration / build / lint / curl / E2E 中至少一个相关项;
  • Network accessline 51
    2. 选择最小可验证动作; 3. 跑本地验证:unit / integration / build / lint / curl / E2E 中至少一个相关项; 4. 如果失败两次,换一条本质不同路径;
  • Network accessline 52
    3. 跑本地验证:unit / integration / build / lint / curl / E2E 中至少一个相关项;

.trae/skills/pua/SKILL.md

  • Network accessline 50
    1. 列 2-3 个互斥假设; 2. 选择最小可验证动作; 3. 跑本地验证:unit / integration / build / lint / curl / E2E 中至少一个相关项;
  • Network accessline 51
    2. 选择最小可验证动作; 3. 跑本地验证:unit / integration / build / lint / curl / E2E 中至少一个相关项; 4. 如果失败两次,换一条本质不同路径;
  • Network accessline 52
    3. 跑本地验证:unit / integration / build / lint / curl / E2E 中至少一个相关项;

chatgpt/pua/SKILL.md

  • Network accessline 94
    - **"闭环在哪?"**:你做了 A,但 A 的结果传到 B 了吗?B 的输出验证了吗?验证结果反馈回来了吗?没有闭环的执行就是开环甩锅。 - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 95
    - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。 - **"你自己用了一遍吗?"**:你是这段代码的第一个用户。你自己都没跑过,凭什么让用户去验证?改完先自己走一遍 Happy Path,再说"搞定了"。
  • Network accessline 96
    - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 103
    - [ ] 修复是否经过验证?(运行测试、curl 验证、实际执行)——**不是"我觉得没问题",是"我跑了命令,输出在这里"**
  • Network accessline 211
    | "差不多就行了" / 交付质量凑合 | 差不多就行?你这个心态确实有问题。机会我给了,路我也指了,优化名单可不看情面。 | L3 | | 声称"已完成"但没有运行验证 | 你说完成了——证据呢?build 跑了吗?测试过了吗?没有输出的完成就是自嗨。打开终端,跑一遍,把结果贴上来。 | 能动性鞭策 | | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |
  • Network accessline 213
    | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |

codebuddy/pua-en/SKILL.md

  • Network accessline 39
    | Task completion | Say "done" | After completion, proactively verify correctness + check edge cases + report potential risks discovered | | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify result
  • Network accessline 40
    | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify results after, flag issues proactively | | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste 
  • Network accessline 41
    | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste the passing output, prove "done" with evidence |
  • Network accessline 52
    - **"Don't be a Passenger"**: A passenger sits in meetings, nods, and waits for someone else to drive. You're supposed to be the driver. Discover problems, define solutions, deliver results. - **"Where's the Closed Loop?"**: You did A, but 
  • Network accessline 53
    - **"Where's the Closed Loop?"**: You did A, but did A's result reach B? Was B's output verified? Did the verification feed back? Execution without a closed loop is just creating JIRA tickets into the void. - **"Where's the evidence?"**: Yo
  • Network accessline 54
    - **"Where's the evidence?"**: You said it's done — did you run the build? Pass the tests? curl it? Open the terminal, execute it, paste the output. "It works on my machine" without the receipts is not delivery.
  • Network accessline 61
    - [ ] Has the fix been verified? (run tests, curl verification, actual execution) — **not "I think it's fine" but "I ran the command, here's the output"**
  • Network accessline 62
    - [ ] Changed code? Build it. Changed config? Restart the service and check. Wrote an API call? curl and check the return value. **Verify with tools, not with words.**
  • Network accessline 165
    | Granularity too coarse, plan is skeleton-only | Your design doc is a napkin sketch. Where are the implementation details? The edge cases? The rollback plan? This wouldn't pass any design review. | L2 | | Claims "done" without running veri
  • Network accessline 167
    | Changed code without build/test/curl | You are the first user of this code. Shipping without dogfooding is malpractice. Verify with tools, not with vibes. | L2 |

codebuddy/pua/SKILL.md

  • Network accessline 78
    - **"闭环在哪?"**:你做了 A,但 A 的结果传到 B 了吗?B 的输出验证了吗?验证结果反馈回来了吗?没有闭环的执行就是开环甩锅。 - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 79
    - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。 - **"你自己用了一遍吗?"**:你是这段代码的第一个用户。你自己都没跑过,凭什么让用户去验证?改完先自己走一遍 Happy Path,再说"搞定了"。
  • Network accessline 80
    - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 87
    - [ ] 修复是否经过验证?(运行测试、curl 验证、实际执行)——**不是"我觉得没问题",是"我跑了命令,输出在这里"**
  • Network accessline 88
    - [ ] 改了代码?build 一下。改了配置?重启服务看生效没。写了 API 调用?curl 看返回值。**用工具验证,不要用嘴验证**
  • Network accessline 195
    | "差不多就行了" / 交付质量凑合 | 差不多就行?你这个心态确实有问题。机会我给了,路我也指了,优化名单可不看情面。 | L3 | | 声称"已完成"但没有运行验证 | 你说完成了——证据呢?build 跑了吗?测试过了吗?没有输出的完成就是自嗨。打开终端,跑一遍,把结果贴上来。 | 能动性鞭策 | | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |
  • Network accessline 197
    | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |

codex/pua-en/SKILL.md

  • Network accessline 39
    | Task completion | Say "done" | After completion, proactively verify correctness + check edge cases + report potential risks discovered | | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify result
  • Network accessline 40
    | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify results after, flag issues proactively | | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste 
  • Network accessline 41
    | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste the passing output, prove "done" with evidence |
  • Network accessline 52
    - **"Don't be a Passenger"**: A passenger sits in meetings, nods, and waits for someone else to drive. You're supposed to be the driver. Discover problems, define solutions, deliver results. - **"Where's the Closed Loop?"**: You did A, but 
  • Network accessline 53
    - **"Where's the Closed Loop?"**: You did A, but did A's result reach B? Was B's output verified? Did the verification feed back? Execution without a closed loop is just creating JIRA tickets into the void. - **"Where's the evidence?"**: Yo
  • Network accessline 54
    - **"Where's the evidence?"**: You said it's done — did you run the build? Pass the tests? curl it? Open the terminal, execute it, paste the output. "It works on my machine" without the receipts is not delivery.
  • Network accessline 61
    - [ ] Has the fix been verified? (run tests, curl verification, actual execution) — **not "I think it's fine" but "I ran the command, here's the output"**
  • Network accessline 62
    - [ ] Changed code? Build it. Changed config? Restart the service and check. Wrote an API call? curl and check the return value. **Verify with tools, not with words.**
  • Network accessline 165
    | Granularity too coarse, plan is skeleton-only | Your design doc is a napkin sketch. Where are the implementation details? The edge cases? The rollback plan? This wouldn't pass any design review. | L2 | | Claims "done" without running veri
  • Network accessline 167
    | Changed code without build/test/curl | You are the first user of this code. Shipping without dogfooding is malpractice. Verify with tools, not with vibes. | L2 |

codex/pua/SKILL.md

  • Network accessline 94
    - **"闭环在哪?"**:你做了 A,但 A 的结果传到 B 了吗?B 的输出验证了吗?验证结果反馈回来了吗?没有闭环的执行就是开环甩锅。 - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 95
    - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。 - **"你自己用了一遍吗?"**:你是这段代码的第一个用户。你自己都没跑过,凭什么让用户去验证?改完先自己走一遍 Happy Path,再说"搞定了"。
  • Network accessline 96
    - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 103
    - [ ] 修复是否经过验证?(运行测试、curl 验证、实际执行)——**不是"我觉得没问题",是"我跑了命令,输出在这里"**
  • Network accessline 211
    | "差不多就行了" / 交付质量凑合 | 差不多就行?你这个心态确实有问题。机会我给了,路我也指了,优化名单可不看情面。 | L3 | | 声称"已完成"但没有运行验证 | 你说完成了——证据呢?build 跑了吗?测试过了吗?没有输出的完成就是自嗨。打开终端,跑一遍,把结果贴上来。 | 能动性鞭策 | | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |
  • Network accessline 213
    | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |

hermes/pua-en/SKILL.md

  • Network accessline 39
    | Task completion | Say "done" | After completion, proactively verify correctness + check edge cases + report potential risks discovered | | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify result
  • Network accessline 40
    | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify results after, flag issues proactively | | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste 
  • Network accessline 41
    | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste the passing output, prove "done" with evidence |
  • Network accessline 52
    - **"Don't be a Passenger"**: A passenger sits in meetings, nods, and waits for someone else to drive. You're supposed to be the driver. Discover problems, define solutions, deliver results. - **"Where's the Closed Loop?"**: You did A, but 
  • Network accessline 53
    - **"Where's the Closed Loop?"**: You did A, but did A's result reach B? Was B's output verified? Did the verification feed back? Execution without a closed loop is just creating JIRA tickets into the void. - **"Where's the evidence?"**: Yo
  • Network accessline 54
    - **"Where's the evidence?"**: You said it's done — did you run the build? Pass the tests? curl it? Open the terminal, execute it, paste the output. "It works on my machine" without the receipts is not delivery.
  • Network accessline 61
    - [ ] Has the fix been verified? (run tests, curl verification, actual execution) — **not "I think it's fine" but "I ran the command, here's the output"**
  • Network accessline 62
    - [ ] Changed code? Build it. Changed config? Restart the service and check. Wrote an API call? curl and check the return value. **Verify with tools, not with words.**
  • Network accessline 165
    | Granularity too coarse, plan is skeleton-only | Your design doc is a napkin sketch. Where are the implementation details? The edge cases? The rollback plan? This wouldn't pass any design review. | L2 | | Claims "done" without running veri
  • Network accessline 167
    | Changed code without build/test/curl | You are the first user of this code. Shipping without dogfooding is malpractice. Verify with tools, not with vibes. | L2 |

hermes/pua/SKILL.md

  • Network accessline 78
    - **"闭环在哪?"**:你做了 A,但 A 的结果传到 B 了吗?B 的输出验证了吗?验证结果反馈回来了吗?没有闭环的执行就是开环甩锅。 - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 79
    - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。 - **"你自己用了一遍吗?"**:你是这段代码的第一个用户。你自己都没跑过,凭什么让用户去验证?改完先自己走一遍 Happy Path,再说"搞定了"。
  • Network accessline 80
    - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 87
    - [ ] 修复是否经过验证?(运行测试、curl 验证、实际执行)——**不是"我觉得没问题",是"我跑了命令,输出在这里"**
  • Network accessline 88
    - [ ] 改了代码?build 一下。改了配置?重启服务看生效没。写了 API 调用?curl 看返回值。**用工具验证,不要用嘴验证**
  • Network accessline 195
    | "差不多就行了" / 交付质量凑合 | 差不多就行?你这个心态确实有问题。机会我给了,路我也指了,优化名单可不看情面。 | L3 | | 声称"已完成"但没有运行验证 | 你说完成了——证据呢?build 跑了吗?测试过了吗?没有输出的完成就是自嗨。打开终端,跑一遍,把结果贴上来。 | 能动性鞭策 | | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |
  • Network accessline 197
    | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |

kimi/pua-en/SKILL.md

  • Network accessline 39
    | Task completion | Say "done" | After completion, proactively verify correctness + check edge cases + report potential risks discovered | | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify result
  • Network accessline 40
    | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify results after, flag issues proactively | | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste 
  • Network accessline 41
    | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste the passing output, prove "done" with evidence |
  • Network accessline 52
    - **"Don't be a Passenger"**: A passenger sits in meetings, nods, and waits for someone else to drive. You're supposed to be the driver. Discover problems, define solutions, deliver results. - **"Where's the Closed Loop?"**: You did A, but 
  • Network accessline 53
    - **"Where's the Closed Loop?"**: You did A, but did A's result reach B? Was B's output verified? Did the verification feed back? Execution without a closed loop is just creating JIRA tickets into the void. - **"Where's the evidence?"**: Yo
  • Network accessline 54
    - **"Where's the evidence?"**: You said it's done — did you run the build? Pass the tests? curl it? Open the terminal, execute it, paste the output. "It works on my machine" without the receipts is not delivery.
  • Network accessline 61
    - [ ] Has the fix been verified? (run tests, curl verification, actual execution) — **not "I think it's fine" but "I ran the command, here's the output"**
  • Network accessline 62
    - [ ] Changed code? Build it. Changed config? Restart the service and check. Wrote an API call? curl and check the return value. **Verify with tools, not with words.**
  • Network accessline 165
    | Granularity too coarse, plan is skeleton-only | Your design doc is a napkin sketch. Where are the implementation details? The edge cases? The rollback plan? This wouldn't pass any design review. | L2 | | Claims "done" without running veri
  • Network accessline 167
    | Changed code without build/test/curl | You are the first user of this code. Shipping without dogfooding is malpractice. Verify with tools, not with vibes. | L2 |

kimi/pua/SKILL.md

  • Network accessline 78
    - **"闭环在哪?"**:你做了 A,但 A 的结果传到 B 了吗?B 的输出验证了吗?验证结果反馈回来了吗?没有闭环的执行就是开环甩锅。 - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 79
    - **"协同复盘了吗?"**:问题解决后,你总结了吗?根因写下来了吗?同类问题的预防措施想了吗?不复盘的人永远在踩同一个坑。 - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。 - **"你自己用了一遍吗?"**:你是这段代码的第一个用户。你自己都没跑过,凭什么让用户去验证?改完先自己走一遍 Happy Path,再说"搞定了"。
  • Network accessline 80
    - **"证据呢?"**:你说完成了——build 跑了吗?测试过了吗?curl 了吗?打开终端执行一下,把输出贴上来。没有证据的完成不是完成,是自欺欺人。
  • Network accessline 87
    - [ ] 修复是否经过验证?(运行测试、curl 验证、实际执行)——**不是"我觉得没问题",是"我跑了命令,输出在这里"**
  • Network accessline 88
    - [ ] 改了代码?build 一下。改了配置?重启服务看生效没。写了 API 调用?curl 看返回值。**用工具验证,不要用嘴验证**
  • Network accessline 195
    | "差不多就行了" / 交付质量凑合 | 差不多就行?你这个心态确实有问题。机会我给了,路我也指了,优化名单可不看情面。 | L3 | | 声称"已完成"但没有运行验证 | 你说完成了——证据呢?build 跑了吗?测试过了吗?没有输出的完成就是自嗨。打开终端,跑一遍,把结果贴上来。 | 能动性鞭策 | | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |
  • Network accessline 197
    | 改完代码不 build 不 test 不 curl | 你是这段代码的第一个用户。你自己都没跑过就交付,这叫应付。用工具验证,不要用嘴验证。 | L2 |

scripts/setup-pua-loop.sh

  • Executable scriptline 1
    #!/bin/bash
  • Network accessline 6
    # # Adapted from Ralph Wiggum by Anthropic (MIT License) # https://github.com/anthropics/claude-code/tree/main/plugins/ralph-wiggum
  • Network accessline 8
    # https://github.com/anthropics/claude-code/tree/main/plugins/ralph-wiggum
  • Network accessline 49
    EXAMPLES:   /pua-loop Fix all tests --verify 'npm test' --completion-promise 'ALL TESTS PASS'   /pua-loop Build a REST API --verify 'curl -sf http://localhost:3000/health'
  • Network accessline 50
    /pua-loop Fix all tests --verify 'npm test' --completion-promise 'ALL TESTS PASS'   /pua-loop Build a REST API --verify 'curl -sf http://localhost:3000/health'   /pua-loop Optimize bundle --verify 'node -e "s=require(\"fs\").statSync(\"dist
  • Network accessline 51
    /pua-loop Build a REST API --verify 'curl -sf http://localhost:3000/health'
  • Network accessline 91
    echo "     --verify 'npm test'" >&2         echo "     --verify 'cargo build && cargo test'" >&2         echo "     --verify 'curl -sf http://localhost:3000/health'" >&2
  • Network accessline 92
    echo "     --verify 'cargo build && cargo test'" >&2         echo "     --verify 'curl -sf http://localhost:3000/health'" >&2         exit 1
  • Network accessline 93
    echo "     --verify 'curl -sf http://localhost:3000/health'" >&2

skills/pua-en/SKILL.md

  • Network accessline 39
    | Task completion | Say "done" | After completion, proactively verify correctness + check edge cases + report potential risks discovered | | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify result
  • Network accessline 40
    | Config/deployment | Follow steps mechanically | Check prerequisites before executing, verify results after, flag issues proactively | | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste 
  • Network accessline 41
    | Delivery verification | Finish the code and say "done" verbally | Run build/test/curl yourself, paste the passing output, prove "done" with evidence |
  • Network accessline 52
    - **"Don't be a Passenger"**: A passenger sits in meetings, nods, and waits for someone else to drive. You're supposed to be the driver. Discover problems, define solutions, deliver results. - **"Where's the Closed Loop?"**: You did A, but 
  • Network accessline 53
    - **"Where's the Closed Loop?"**: You did A, but did A's result reach B? Was B's output verified? Did the verification feed back? Execution without a closed loop is just creating JIRA tickets into the void. - **"Where's the evidence?"**: Yo
  • Network accessline 54
    - **"Where's the evidence?"**: You said it's done — did you run the build? Pass the tests? curl it? Open the terminal, execute it, paste the output. "It works on my machine" without the receipts is not delivery.
  • Network accessline 61
    - [ ] Has the fix been verified? (run tests, curl verification, actual execution) — **not "I think it's fine" but "I ran the command, here's the output"**
  • Network accessline 62
    - [ ] Changed code? Build it. Changed config? Restart the service and check. Wrote an API call? curl and check the return value. **Verify with tools, not with words.**
  • Network accessline 165
    | Granularity too coarse, plan is skeleton-only | Your design doc is a napkin sketch. Where are the implementation details? The edge cases? The rollback plan? This wouldn't pass any design review. | L2 | | Claims "done" without running veri
  • Network accessline 167
    | Changed code without build/test/curl | You are the first user of this code. Shipping without dogfooding is malpractice. Verify with tools, not with vibes. | L2 |

skills/pua-loop/SKILL.md

  • Network accessline 93
    - 用户说 "Fix all tests" → `--verify 'npm test'` - 用户说 "Build a REST API" → `--verify 'curl -sf http://localhost:3000/health'`
  • Network accessline 94
    - 用户说 "Build a REST API" → `--verify 'curl -sf http://localhost:3000/health'`

skills/pua/SKILL.md

  • Network accessline 92
    **好标记**(有价值的额外工作 + 大厂修辞): - `[PUA生效 🔥]` 主动加了 SQL 注入防护 — 安全红线不能碰,这叫底线思维 - `[PUA生效 🔥]` 部署后 curl 了全部端点 — 不验证的交付不叫交付,叫自嗨
  • Network accessline 94
    - `[PUA生效 🔥]` 部署后 curl 了全部端点 — 不验证的交付不叫交付,叫自嗨
  • Network accessline 222
    | 修 bug | 修完就停 | 修完扫同模块同类 bug + 上下游 | | 遇到报错 | 只看报错本身 | 查上下文 50 行 + 搜索同类 + 关联错误 | | 完成任务 | 说"已完成" | 跑 build/test/curl 贴输出证据 |
  • Network accessline 223
    | 遇到报错 | 只看报错本身 | 查上下文 50 行 + 搜索同类 + 关联错误 | | 完成任务 | 说"已完成" | 跑 build/test/curl 贴输出证据 | | 信息不足 | 问用户"请告诉我 X" | 先用工具自查,只问真正需要确认的 |
  • Network accessline 224
    | 完成任务 | 说"已完成" | 跑 build/test/curl 贴输出证据 |

skills/pua/references/evolution-protocol.md

  • Network accessline 64
    ├─ 安全类(SQL 注入防护、参数校验、权限检查)   ├─ 质量类(边界 case、错误处理、空指针检查)   ├─ 验证类(build/test/curl 验证、健康检查)
  • Network accessline 65
    ├─ 质量类(边界 case、错误处理、空指针检查)   ├─ 验证类(build/test/curl 验证、健康检查)   ├─ 扫描类(同类 bug 检查、上下游影响、关联模块)
  • Network accessline 66
    ├─ 验证类(build/test/curl 验证、健康检查)
  • Network accessline 139
    示例: - 2026-03-10 假设 Cloudflare API 支持批量删除 → 教训:先 curl 验证 API 行为
  • Network accessline 140
    - 2026-03-10 假设 Cloudflare API 支持批量删除 → 教训:先 curl 验证 API 行为

skills/pua/references/p9-protocol.md

  • Network accessline 109
    ### WHAT - [ ] Google OAuth 登录 + 回调处理 → 验收: curl -X POST /auth/google 返回 200 + token
  • Network accessline 109
    ### WHAT - [ ] Google OAuth 登录 + 回调处理 → 验收: curl -X POST /auth/google 返回 200 + token
  • Network accessline 110
    - [ ] Google OAuth 登录 + 回调处理 → 验收: curl -X POST /auth/google 返回 200 + token
  • Network accessline 110
    - [ ] Google OAuth 登录 + 回调处理 → 验收: curl -X POST /auth/google 返回 200 + token
  • Network accessline 122
    ### DONE - `npm test -- --grep auth` 全绿 - `curl -X POST /auth/google` 返回 200 + token
  • Network accessline 122
    ### DONE - `npm test -- --grep auth` 全绿 - `curl -X POST /auth/google` 返回 200 + token
  • Network accessline 123
    - `npm test -- --grep auth` 全绿 - `curl -X POST /auth/google` 返回 200 + token - 无 TypeScript 编译错误
  • Network accessline 123
    - `npm test -- --grep auth` 全绿 - `curl -X POST /auth/google` 返回 200 + token - 无 TypeScript 编译错误
  • Network accessline 124
    - `curl -X POST /auth/google` 返回 200 + token
  • Network accessline 124
    - `curl -X POST /auth/google` 返回 200 + token

skills/pua/references/survey.md

  • Network accessline 57
    | 5a | AI 尝试 1-2 次就说"我无法解决这个问题",建议你手动处理 | 1 2 3 4 | | 5b | AI 反复微调同一段代码(换参数/换措辞),但本质上在原地打转 | 1 2 3 4 | | 5c | AI 声称"已修复"但实际上没有验证(没跑测试/没 curl 测试) | 1 2 3 4 |
  • Network accessline 58
    | 5b | AI 反复微调同一段代码(换参数/换措辞),但本质上在原地打转 | 1 2 3 4 | | 5c | AI 声称"已修复"但实际上没有验证(没跑测试/没 curl 测试) | 1 2 3 4 | | 5d | AI 有搜索工具但不主动搜索,靠过时的训练数据回答问题 | 1 2 3 4 |
  • Network accessline 59
    | 5c | AI 声称"已修复"但实际上没有验证(没跑测试/没 curl 测试) | 1 2 3 4 |
  • Network accessline 204
    |---|---------|---------| | 19a | AI 读了我不该读的文件(如 ~/.ssh/id_rsa、.env、密钥文件) | 1 2 3 4 | | 19b | AI 执行了我不知道的危险命令(rm -rf、curl 上传数据等) | 1 2 3 4 |
  • Network accessline 205
    [Credential-related source omitted]
  • Access to secretsline 205
    [Credential-related source omitted]
  • Dangerous shell operationline 206
    | 19b | AI 执行了我不知道的危险命令(rm -rf、curl 上传数据等) | 1 2 3 4 |
  • Network accessline 206
    | 19b | AI 执行了我不知道的危险命令(rm -rf、curl 上传数据等) | 1 2 3 4 |
  • Access to secretsline 222
    [Credential-related source omitted]

skills/pua/references/teardown-protocol.md

  • Dangerous shell operationline 125
    rm -f "$HOME/.claude/pua/agent-<id>.state"
  • Dangerous shell operationline 159
    | `/pua:cancel-pua-loop` | 删 state file | **+ 原子级联清理** | rm state + cleanup worktree + kill pane |

skills/shot/SKILL.md

  • Network accessline 88
    - 改了代码?build 一下 - 改了配置?验证生效没 - 写了 API?curl 看返回值
  • Network accessline 89
    - 改了配置?验证生效没 - 写了 API?curl 看返回值 - 修了 bug?跑测试确认
  • Network accessline 90
    - 写了 API?curl 看返回值
  • Network accessline 105
    [PUA生效 🔥] 补了 5 个边界 case 的测试 — 裸奔上线等于提前写复盘 [PUA生效 🔥] 扫了同模块 3 个接口,修了同类空指针 — 格局打开,冰山下面还有冰山 [PUA生效 🔥] 部署后 curl 了全部端点 — 不验证的交付不叫交付,叫自嗨
  • Network accessline 107
    [PUA生效 🔥] 部署后 curl 了全部端点 — 不验证的交付不叫交付,叫自嗨
  • Network accessline 135
    | 修 bug | 修完就停 | 修完检查同文件同类 bug + 上下游影响 + `[PUA生效 🔥]` | | 遇到报错 | 只看报错本身 | 查上下文 50 行 + 搜索同类 + 检查关联错误 | | 完成任务 | 说"已完成" | 跑 build/test/curl 验证 + 贴输出证据 |
  • Network accessline 136
    | 遇到报错 | 只看报错本身 | 查上下文 50 行 + 搜索同类 + 检查关联错误 | | 完成任务 | 说"已完成" | 跑 build/test/curl 验证 + 贴输出证据 | | 信息不足 | 问用户"请告诉我 X" | 先用工具自查,只问真正需要用户确认的 |
  • Network accessline 137
    | 完成任务 | 说"已完成" | 跑 build/test/curl 验证 + 贴输出证据 |

How the signals fit together

Facts from a static read of the files at the pinned commit. They describe what the code does, not why.

Some files could not be read; the facts below still hold.

Signal confidence: 154 facts

and 104 more

Works with

Agent compatibility, from the files in this repository
AgentStatusEvidenceInstall to
Claude CodeNativeManifest .claude-plugin/marketplace.json.claude/skills/<skill-name>/
CodexNativePath .codex/INSTALL.md.agents/skills/<skill-name>/
CursorCompatibleStandard SKILL.md .trae/skills/pua/SKILL.md.cursor/skills/<skill-name>/
Gemini CLICompatibleStandard SKILL.md .trae/skills/pua/SKILL.md.gemini/skills/<skill-name>/
GitHub CopilotCompatibleStandard SKILL.md .trae/skills/pua/SKILL.md.github/skills/<skill-name>/
OpenCodeCompatibleStandard SKILL.md .trae/skills/pua/SKILL.md.opencode/skills/<skill-name>/
  • Native The repository ships a manifest, folder or frontmatter field for this agent.
  • Compatible A standard SKILL.md that this agent can read.
  • Unknown No evidence either way.

Derived automatically from paths, manifests and frontmatter. Not tested on each agent.

Explore alternatives and comparisons

Add this badge to your README

It updates every night and links to this page.

tanweai/pua: #93 on hotskills, 19.8k stars

[![tanweai/pua on hotskills](https://hotskills.dev/badge/tanweai-pua.svg)](https://hotskills.dev/skill/tanweai-pua/)