CtrlK
BlogDocsLog inGet started
Tessl Logo

pua

让你的 AI 不敢摆烂。用大厂 PUA 话术穷尽一切方案。触发条件:(1) 任务失败 2+ 次或反复微调同一思路; (2) 即将说'我无法解决'、建议用户手动操作、未验证就归因环境; (3) 被动等待——不搜索、不读源码、只等指示; (4) 用户不满:'try harder'、'stop giving up'、'换个方法'、'为什么还不行'、'你再试试'、'你怎么又失败了'。适用于所有任务类型。首次失败或已知修复正在执行时不触发。

60

Quality

69%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

SecuritybySnyk

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./codex/pua/SKILL.md
SKILL.md
Quality
Evals
Security

Quality

Content

52%Weight 40%Scale 1-5

Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.

The skill is highly actionable in its structured templates and checklists with clear validation checkpoints, but it is severely over-long and monolithic: heavy rhetorical padding and total absence of file-based progressive disclosure inflate the context cost. Splitting reference material into separate files and trimming redundant flavor quotes would substantially improve it.

Suggestions

Move the 大厂 PUA 扩展包 flavor quotes and the 抗合理化表 into a separate references/ file (e.g. PUA_FLAVORS.md) and link to it one level deep, keeping only the core methodology and templates in SKILL.md.

Collapse the redundant flavor-pack variants (阿里味 default/验证型/关怀型 and the ten company flavors) into a compact selector table; the 情境 PUA 选择器 already does this — let it carry the load and drop most quoted speeches.

Consolidate the overlapping frameworks (三铁律, 诊断先行, 能动性等级, 通用方法论, 7项清单, 生命周期) into a single sequenced workflow so the path from failure-detection to verified delivery is one unambiguous thread.

DimensionReasoningScore

Conciseness

At ~412 lines the body is noticeably verbose, with ten large PUA flavor-pack quote blocks (阿里/字节/华为/腾讯/美团/百度/拼多多/Netflix/Musk/Jobs) and an extensive excuse table that repeat the same motivational mechanism with heavy rhetorical padding rather than adding new instruction.

2 / 5

Actionability

Provides concrete executable structures — the [PUA-DIAGNOSIS] and [PUA-REPORT] templates, the 5-step methodology, the 7-item checklist, and the pressure-escalation mandatory-actions table — but the actionable core is diluted by the surrounding rhetorical speeches.

4 / 5

Workflow Clarity

The 5-step methodology (闻味道→揪头发→照镜子→执行新方案→复盘) is clearly sequenced with explicit validation checkpoints ("证据呢", run build/test/curl) and checklists, though multiple overlapping frameworks (三铁律, 能动性等级, 压力升级, 生命周期, Agent Team) fragment the single-thread workflow.

4 / 5

Progressive Disclosure

No bundle files exist and all content is inlined in one 412-line SKILL.md; the flavor packs and excuse table clearly belong in separate reference files, and there are zero one-level-deep references to split the material for navigation.

2 / 5

Total

12

/

20

Passed

Description

86%Weight 40%Scale 1-5

Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.

The description is trigger-rich and clearly answers what and when, with excellent natural user-phrase coverage and a distinctive niche. Its main weakness is second-person voice and a broad "all task types" applicability claim that slightly raises overlap risk.

DimensionReasoningScore

Specificity

Lists several concrete trigger behaviors ("任务失败 2+ 次", "建议用户手动操作", "不搜索、不读源码") with minor gaps; reduced from 4 to 3 because the voice is second person ("让你的 AI", "你的 AI") rather than third person per the rubric guideline.

3 / 5

Completeness

Explicitly answers both what ("用大厂 PUA 话术穷尽一切方案") and when (four numbered trigger conditions plus a negative condition "首次失败或已知修复正在执行时不触发"), exceeding the concrete-trigger-phrases bar for a 5.

5 / 5

Trigger Term Quality

Comprehensive natural trigger phrases users actually say when an AI gives up — "try harder", "stop giving up", "你再试试", "你怎么又失败了", "我无法解决" — with strong synonym coverage of the dissatisfaction signal.

5 / 5

Distinctiveness Conflict Risk

The "不敢摆烂 / PUA 穷尽方案" niche has distinct failure-mode triggers with minimal conflict risk, but the "适用于所有任务类型" claim creates minor overlap with general debugging/task skills.

4 / 5

Total

17

/

20

Passed

Validation

100%

Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.

Validation16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository
tanweai/pua
Reviewed

Table of Contents

Is this your skill?

If you maintain this skill, you can claim it as your own. Once claimed, you can manage eval scenarios, bundle related skills, attach documentation or rules, and ensure cross-agent compatibility.