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

88%

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

77%

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

A well-structured, actionable wrapper skill with a clear validated workflow; its main weaknesses are some over-explanatory prose and a bundle that omits the vendored upstream/ tree the body and detector both depend on.

Suggestions

Trim the `technical`-mode and CLAUDE_PLUGIN_ROOT/Codex asides to the essential one-liner each; Claude does not need the context-vs-voice distinction spelled out at length.

Ship the referenced `upstream/` directory (upstream/SKILL.md, upstream/UPSTREAM-PIN, upstream/detector/patterns.js) or note explicitly that detect.js is unavailable until the plugin fetches it, since detect.js fails with MODULE_NOT_FOUND without it.

If upstream/ is intentionally not bundled, replace the unconditional "run the bundled detector" instruction with a guarded step so the workflow does not dead-end on a missing dependency.

Dimension	Reasoning	Score
Conciseness	Mostly efficient with executable commands and a tight wrapper framing, but includes some over-explanation Claude doesn't need — the `technical`-mode digression ("It does not broadly relax vocabulary checks... context and voice are separate axes") and the CLAUDE_PLUGIN_ROOT-vs-Codex aside could be trimmed.	2 / 3
Actionability	Gives copy-paste-ready commands ("node \"$SKILL_DIR/scripts/detect.js\" FILE"), lists valid modes, and describes the JSON output fields — fully executable, not pseudocode.	3 / 3
Workflow Clarity	A numbered run sequence with an explicit validation/convergence feedback loop ("repeat the audit -> revise cycle until a scored result has no issues, or the cap of 2 passes") plus the UNSCORED edge-case checkpoint, matching the top anchor.	3 / 3
Progressive Disclosure	Structurally a concise overview pointing one level deep to upstream/SKILL.md and the detector, but the referenced upstream/ directory is absent from the bundle (only scripts/detect.js ships, and it requires the missing upstream/detector/patterns.js), so the signaled references do not fully resolve.	2 / 3
	Total	10 / 12 Passed

Description

100%

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, third-person description that specifies concrete capabilities, natural trigger terms, and explicit use-conditions while drawing a clear lane against a sibling skill. It hits the top anchor on every dimension.

Dimension	Reasoning	Score
Specificity	Names multiple concrete actions — "Remove AI-writing patterns", "detect-only audit, edit-in-place, voice profiles (casual / professional / technical / warm / blunt), and iterate-to-convergence" — matching the multi-action anchor.	3 / 3
Completeness	Explicitly answers both what (remove AI-isms via the catalog) and when ("Use ONLY when the user explicitly asks...") with concrete triggers, satisfying the top anchor.	3 / 3
Trigger Term Quality	Quotes natural user phrasings users would actually say — "remove AI patterns / AI-isms", "make this sound less like AI", "de-AI this text" — giving good coverage, not jargon.	3 / 3
Distinctiveness Conflict Risk	Has a clear niche plus an explicit carve-out — "NOT for clarity/composition rules or formal technical-prose polish — use agent-style for that" and "Do not auto-trigger" — making wrong-skill conflicts unlikely.	3 / 3
	Total	12 / 12 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: 45ee0e2

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.