cw-ai-check

Scan or clean prose for AI-generated residue, including unsupported connective tissue, epistemic inflation, over-completed arguments, performed voice, formulaic structure, empty figurative language, stock phrasing, corporate abstraction, hedging, and false enthusiasm. Use for AI-smell cleanup, humanizing model-assisted prose, source-faithful drafting validation, writing polish, or pre-publication checks; enforce silently when composed into drafting or line editing.

Quality

86%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Quality

Content

86%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.

A well-structured instruction-only skill with concrete per-category flag/fix/protect guidance, a sequenced scan order, and clean progressive disclosure to two real reference files. The main improvement opportunity is making validation checkpoints explicit in the cleanup workflow.

Suggestions

Add an explicit validate-fix-retry checkpoint to the Manual Cleanup workflow (e.g., re-scan the cleaned copy against the lexicon before returning it, and re-run §11 self-commentary detection on the revision).

Tighten the densest enumeration sections (e.g., Categories 2 and 7) by grouping near-synonyms or pruning redundant example phrases now that the full lexicon lives in the CSV reference.

Cross-link each Detection Category that defers detail to structural-overcompletion.md (notably §11/§12) with an explicit pointer line so the deferral is visible at the category, not just in the opening note.

Dimension	Reasoning	Score
Conciseness	Mostly efficient and assumes Claude's competence (no primers on what AI or metaphors are), with the heavy lexicon correctly offloaded to a reference file; a few dense enumeration sections could still be tightened.	4 / 5
Actionability	Each detection category supplies concrete flagged example phrases, a specific Fix directive, and Protect exceptions, and the output format templates are copy-paste ready, giving fully executable instruction-only guidance.	5 / 5
Workflow Clarity	The seven-pass Scan Order is explicitly sequenced with conditional handling and a stated exception, but validation checkpoints are implicit rather than explicit validate-fix-retry gates.	4 / 5
Progressive Disclosure	The body is a clear overview with two well-signaled, one-level-deep references (structural-overcompletion.md and ai_tells_lexicon.csv, both confirmed present), and the bulk 293-row lexicon is appropriately split into its own file.	5 / 5
	Total	18 / 20 Passed

Description

87%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.

A strong description that clearly states both capability and trigger conditions with concrete, natural-language phrasing. It distinguishes the skill well and lists numerous specific detection targets; only minor keyword-synonym and action-variety gaps keep it just below a perfect specificity score.

Dimension	Reasoning	Score
Specificity	Names concrete actions ("Scan or clean prose for AI-generated residue") and enumerates many specific tell-types ("unsupported connective tissue, epistemic inflation, over-completed arguments, performed voice"), giving several specific detection targets with only minor coverage gaps.	4 / 5
Completeness	Explicitly answers both what ("Scan or clean prose for AI-generated residue, including...") and when ("Use for AI-smell cleanup, humanizing model-assisted prose...") with concrete trigger phrases.	5 / 5
Trigger Term Quality	Includes natural user phrases ("AI-smell cleanup," "humanizing model-assisted prose," "writing polish," "pre-publication checks") that users would actually say, though a few synonyms like "de-AI" or "robotic prose" are missing.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (AI-residue detection and removal in prose) with distinct triggers unlikely to fire for the wrong skill, presenting minimal conflict risk.	5 / 5
	Total	18 / 20 Passed

Validation

100%

Warnings & errors only

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

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository: EveryInc/compound-writing
Path: skills/cw-ai-check/SKILL.md
Commit: 18702f0

Reviewed: 2 days ago

Table of Contents

Discovery Implementation Validation

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.