remove-ai-patterns

Remove AI-writing patterns ("AI-isms") from text using the avoid-ai-writing catalog (conorbronsdon/avoid-ai-writing). Use ONLY when the user explicitly asks to remove AI patterns / AI-isms, "make this sound less like AI", "de-AI this text", or invokes remove-ai-patterns by name. Supports detect-only audit, edit-in-place, voice profiles (casual / professional / technical / warm / blunt), and iterate-to-convergence. NOT for clarity/composition rules or formal technical-prose polish — use agent-style for that. Do not auto-trigger for ordinary writing or editing tasks.

Quality

94%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Low

Low-risk findings worth noting

Quality

Content

88%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 strong, highly actionable body: executable commands with path-resolution traps preempted, an explicit validation and convergence loop, and clean delegation to the upstream catalog. The main deductions are over-explained asides (SKILL_DIR resolution, the technical-mode digression) and bundle references that cannot be verified against the provided file listing.

Suggestions

Tighten the SKILL_DIR section: drop the conversational aside ('You already know this skill's directory…') and the confusing 'plugins/writing segment' parenthetical, keeping just the two one-line resolution examples.

Compress the technical-mode explanation to its one concrete effect (skips title-case-header) plus a single sentence noting it is distinct from the technical voice profile.

Ensure the referenced bundle files (upstream/SKILL.md, upstream/detector/validate.js, upstream/UPSTREAM-PIN, scripts/update-upstream.sh) actually ship with the skill so every referenced path resolves.

Dimension	Reasoning	Score
Conciseness	Mostly efficient — the MODULE_NOT_FOUND path trap and the UNSCORED edge-case rule are non-obvious operational detail — but the SKILL_DIR resolution aside ('You already know this skill's directory…', the 'plugins/writing segment' parenthetical) and the five-line digression on what the technical context mode does and does not do could be trimmed, fitting the 4 anchor rather than 5.	4 / 5
Actionability	Fully executable, copy-paste-ready commands ('node "$SKILL_DIR/scripts/detect.js" FILE', the validator invocation with absolute path), enumerated valid modes, described JSON output fields, and a concrete convergence rule with an explicit 2-pass cap — the common cases are covered.	5 / 5
Workflow Clarity	A clearly numbered 1–4 sequence with explicit validation checkpoints (scored detector audit, validator where 'a failure [is] a blocking problem', the scored-not-UNSCORED guard) and a feedback loop (audit → revise → re-audit, capped at 2 passes) with anti-early-stop guidance.	5 / 5
Progressive Disclosure	The body is a well-signaled overview delegating the catalog one level deep to 'upstream/SKILL.md', but scored against the actual bundle only 'scripts/detect.js' is present — 'upstream/SKILL.md', 'upstream/detector/validate.js', 'upstream/UPSTREAM-PIN', and 'scripts/update-upstream.sh' are referenced but not verifiable in the provided bundle, a minor organization gap.	4 / 5
	Total	18 / 20 Passed

Description

100%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.

An exemplary description: concrete capabilities, verbatim user trigger phrasings, explicit use-only-when guidance, and a negative boundary that redirects the nearest competing skill. It fully satisfies the what/when/trigger/distinctiveness requirements with no padding.

Dimension	Reasoning	Score
Specificity	Lists multiple concrete capabilities — 'detect-only audit, edit-in-place, voice profiles (casual / professional / technical / warm / blunt), and iterate-to-convergence' — comprehensively covering the skill's modes with no gaps, matching the top anchor rather than the 4 anchor.	5 / 5
Completeness	Explicitly answers both 'what' (catalog-based AI-pattern removal with its four modes) and 'when' ('Use ONLY when the user explicitly asks to remove AI patterns / AI-isms… or invokes remove-ai-patterns by name') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Quotes verbatim natural phrasings users would say — 'remove AI patterns / AI-isms', 'make this sound less like AI', 'de-AI this text' — plus the skill name, giving comprehensive synonym coverage; not the 4 anchor since nothing material is missing.	5 / 5
Distinctiveness Conflict Risk	Clear niche with an explicit negative boundary — 'NOT for clarity/composition rules or formal technical-prose polish — use agent-style for that. Do not auto-trigger for ordinary writing or editing tasks' — minimizing conflict with the overlapping sibling skill.	5 / 5
	Total	20 / 20 Passed

Validation

93%

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 — 15 / 16 Passed

Validation for skill structure

Criteria	Description	Result
referenced_paths_exist	Referenced path issues: 1 missing	Warning

	Total	15 / 16 Passed

Repository: pchalasani/claude-code-tools
Path: plugins/writing/skills/remove-ai-patterns/SKILL.md
Commit: 0ca7320

Reviewed: 14 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.