CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

nopua

The anti-PUA. Drives AI with wisdom, trust, and inner motivation instead of fear and threats. Activates on: task failed 2+ times, about to give up, suggesting user do it manually, blaming environment unverified, stuck in loops, passive behavior, or user frustration ('try harder', 'figure it out', '换个方法', '为什么还不行'). ALL task types. Not for first failures.

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/nopua/SKILL.md

The canonical home for this skill is nopua in wuji-labs/nopua

SKILL.md
Quality
Evals
Security

Quality

Content

56%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 body delivers genuinely actionable behavioral guidance — concrete thresholds, checklists, and report templates — organized in clear sections. However, it is heavily padded with repeated directives and philosophical quotation, and its 450 lines are entirely inlined where a lean overview plus separate reference files would serve better.

Suggestions

Consolidate the repeated verification guidance into one authoritative section: 'tool-verify, don't mouth-verify', the 50-lines-of-context rule, and the delivery checklist currently each appear 3-4 times across the Beliefs, Proactivity Spectrum, Forge, and Practice sections.

Move the Seven Ways and Agent Team Integration into separate reference files (e.g. references/seven-ways.md, references/team-integration.md), keeping a short selector table in SKILL.md that points to them, and remove or create the missing 'agents/nopua-mentor.md' reference.

Trim the Dao De Jing block quotes to single-line epigraphs or remove them; they contribute substantial tokens without changing model behavior.

DimensionReasoningScore

Conciseness

The ~450-line body repeats the same guidance many times — 'tool-verify, don't mouth-verify' and the verification checklist appear in the Beliefs, Proactivity Spectrum, Delivery Checklist, Way of the Forge, and Way of Practice sections, with extended Dao De Jing quotations as padding. This is noticeably verbose (anchor 2) but not anchor 1, since it doesn't explain basic concepts Claude lacks and its tables are information-dense.

2 / 5

Actionability

Guidance is mostly concrete and executable: 'Checks 50 lines of context + searches similar issues', 'search the complete error message', failure-count thresholds mapped to named levels, a 7-item checklist, and copy-paste templates ([NOPUA-REPORT] format, structured handoff format). It falls short of 5 because some directives remain abstract, e.g. 'propose a hypothesis 180° opposite to your current direction' without operational steps.

4 / 5

Workflow Clarity

The Water Methodology gives a clearly sequenced 5-step process (Stop → Observe → Turn → Act → Realize) with per-step detail, and the Delivery Checklist and 7-Point Clarity Checklist supply verification checkpoints. Not 5 because three overlapping control flows (Cognitive Elevation table, Situation Wisdom Selector, Seven Ways) are never reconciled — it's ambiguous which governs when.

4 / 5

Progressive Disclosure

Section headers and tables provide reasonable navigation, but this is a monolithic 450-line file with no bundle structure; the Seven Ways (seven multi-paragraph philosophical essays) and Agent Team Integration clearly belong in separate reference files. It also references 'agents/nopua-mentor.md', which does not exist in the bundle. Fits anchor 3: some structure, but content that should be separate is inline.

3 / 5

Total

13

/

20

Passed

Description

68%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 has an excellent, explicit trigger section with natural multilingual trigger phrases and a clear exclusion ('Not for first failures'), but the capability statement is abstract motivational language rather than concrete actions. Distinctiveness is diluted by the 'ALL task types' scope declaration.

DimensionReasoningScore

Specificity

The capability statement 'Drives AI with wisdom, trust, and inner motivation instead of fear and threats' is abstract motivational language with no concrete actions named, matching the anchor 'names the domain but actions are minimal or generic'. It scores above 1 only because the behavioral domain (anti-PUA / failure-persistence steering) is identifiable.

2 / 5

Completeness

Both what ('Drives AI with wisdom, trust, and inner motivation instead of fear and threats') and when ('Activates on: ...', 'Not for first failures') are explicitly present with concrete triggers, but the 'what' is vague about what the skill actually does. It fits anchor 4 (both present, one could be more specific) rather than 5 (both concrete).

4 / 5

Trigger Term Quality

Triggers are comprehensive and natural: 'task failed 2+ times', 'about to give up', 'suggesting user do it manually', 'stuck in loops', 'passive behavior', plus quoted user phrases 'try harder', 'figure it out', '换个方法', '为什么还不行' — covering behavioral signals, synonyms, and both English and Chinese user phrasings. Not below 5 because no common variation of the trigger scenario is missing.

5 / 5

Distinctiveness Conflict Risk

The failure-escalation triggers ('task failed 2+ times', 'about to give up') are fairly distinct, but 'ALL task types' plus 'passive behavior' creates overlap risk with general debugging, verification, and proactivity skills. This sits between anchor 3 (could still overlap with similar skills) and anchor 4, but the declared universal scope keeps it at 3.

3 / 5

Total

14

/

20

Passed

Validation

93%

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

Validation — 15 / 16 Passed

Validation for skill structure

CriteriaDescriptionResult

frontmatter_unknown_keys

Unknown frontmatter key(s) found; consider removing or moving to metadata

Warning

Total

15

/

16

Passed

Repository

wuji-labs/nopua

Path

codex/nopua/SKILL.md

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.