CtrlK
BlogDocsLog inGet started
Tessl Logo
SkillPublic

jev-prose

Check prose with Jev for narrow AI-writing patterns and revise from its feedback. Use when asked for Jev prose checks, AI-pattern cleanup, or an ongoing check-and-rewrite loop while writing sentences or paragraphs.

Quality

87%

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

The canonical home for this skill is jev-prose in pchalasani/claude-code-tools

SKILL.md
Quality
Evals
Security

Quality

Content

85%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.

Excellent procedural skill: an executable, bounded check-revise loop with strong validation checkpoints and honest failure handling. The only tightening opportunities are a slightly redundant policy sentence and the unspecified report.json findings format.

Suggestions

Document the shape of findings in report.json (field names for flagged questions and probability) so step 4's 'inspect the flagged questions' is fully executable without runtime discovery.

Merge the redundant claim-preservation sentences in the closing policy paragraphs into one statement.

DimensionReasoningScore

Conciseness

The body is dense and directive with no padding or explanations of known concepts, but a few sentences restate an already-implied policy (e.g., 'An unsupported claim is not permission to silently replace the author's intended claim' partially repeats the preceding preservation rule), so it falls just short of lean.

4 / 5

Actionability

Copy-paste-ready commands ('jev-prose check draft.txt --profile general > report.json', 'uv tool install ./packages/jev-prose'), concrete flags, exit-code semantics, and report fields to verify ('ran: true', 'ok: true', 'answered == expected'). Minor gap: the structure of findings inside report.json beyond those verification fields is unspecified, leaving 'inspect the flagged questions' to runtime inspection.

4 / 5

Workflow Clarity

A five-step loop with explicit validation checkpoints (exit-code meanings, ran/ok/answered == expected, exit 2 stops the loop), a bounded retry of three revisions and four checks, a never-end-on-unchecked-rewrite rule, and early-stop conditions for feedback cycling — a textbook validate-fix-retry sequence.

5 / 5

Progressive Disclosure

No bundle files exist and the body references none, so nothing points to missing files. The ~66-line body is well-sectioned (Setup, Bounded loop, closing policies), nothing belongs in a separate file, and single-file organization is appropriate for this scope.

5 / 5

Total

18

/

20

Passed

Description

82%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: third-person, concrete, with an explicit 'Use when' trigger clause and a clearly distinct niche. Its only weaknesses are the thin action list (check + revise only) and missing common synonyms a user might say.

Suggestions

Add one or two more concrete actions to the what-clause (e.g., 'flag AI-sounding passages and report unresolved findings') to lift specificity past two actions.

Include natural user synonyms such as 'make writing sound less AI' or 'humanize prose' alongside 'AI-pattern cleanup' to broaden trigger coverage.

DimensionReasoningScore

Specificity

Names the domain ('narrow AI-writing patterns') and exactly two concrete actions — 'Check prose' and 'revise from its feedback' — matching the 1-2-actions anchor rather than the several-actions anchor at 4; no coverage of reporting, dismissing flags, or other loop behaviors.

3 / 5

Completeness

Explicitly answers both: the what ('Check prose with Jev for narrow AI-writing patterns and revise from its feedback') and a concrete when ('Use when asked for Jev prose checks, AI-pattern cleanup, or an ongoing check-and-rewrite loop...'), matching the top anchor exactly.

5 / 5

Trigger Term Quality

'Jev prose checks', 'AI-pattern cleanup', and 'check-and-rewrite loop' are natural user phrasings, but common synonyms like 'make it sound less AI' or 'humanize the writing' are missing, so coverage is good rather than comprehensive.

4 / 5

Distinctiveness Conflict Risk

'Jev' is a unique tool name and the AI-writing-pattern niche has distinct, specific triggers, giving minimal conflict risk with any other skill.

5 / 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.

Validation — 16 / 16 Passed

Validation for skill structure

No warnings or errors.

Repository

pchalasani/claude-code-tools

Path

packages/jev-prose/src/jev_prose/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.