cowart-image-edit

Generate new AI images from user-supplied Cowart annotation screenshots. Use when the user provides one or more screenshots showing Cowart images marked with the 批注 tool, arrows, or visible edit notes and wants Codex to apply those requested changes, create revised bitmap images, and place each result beside the corresponding original or in a nearby clear area without replacing, moving, hiding, or deleting the original images or annotations.

Quality

79%

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

Fix and improve this skill with Tessl

tessl review fix ./skills/cowart-image-edit/SKILL.md

Quality

Content

71%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 highly actionable, well-sequenced operational skill with concrete MCP call shapes and explicit verification. Its main weaknesses are redundant guardrails that repeat the workflow and a monolithic single-file structure that could benefit from reference files for detailed placement/fallback specs.

Suggestions

De-duplicate the Guardrails section against the workflow: keep only the guardrails that add new constraints not already stated in steps 5–7, or reference the steps instead of restating them.

Extract the detailed placement rules and fallback storage-path specifications into a references/ file (e.g. PLACEMENT.md) and link to it from step 6/7 to reduce SKILL.md length and improve progressive disclosure.

Add an explicit error-recovery loop after the step 8 'Verify visually' checklist (e.g. 'if any check fails, regenerate or re-place the image and re-verify') to reach full workflow-clarity maturity.

Dimension	Reasoning	Score
Conciseness	The body is mostly efficient operational guidance, but the Guardrails section substantially reiterates workflow rules ('Never replace the original image', 'Never delete or move annotation shapes', 'Never auto-capture or scan the current canvas') and the fallback storage-path detail could be tightened.	3 / 5
Actionability	Provides fully executable, copy-paste-ready guidance: the exact MCP tool name, a timestamped filename pattern, a concrete JSON call shape, a specific margin of 40 canvas units, and explicit fallback paths covering the common cases.	5 / 5
Workflow Clarity	An 8-step sequence with explicit validation checkpoints (step 5 visual inspection with resolution-order fallbacks, step 8 'Verify visually' checklist) and ambiguity-handling; falls just short of 5 because the verify step lacks an explicit 'if verification fails, fix and retry' recovery loop.	4 / 5
Progressive Disclosure	Content is organized into clear sections (Preconditions, Workflow, Guardrails) but everything is inlined in a single ~160-line SKILL.md with no one-level-deep reference files for the detailed placement rules and fallback storage specs that could live separately.	3 / 5
	Total	15 / 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, specific description that clearly states both capability and trigger conditions with concrete natural-language terms. It is somewhat verbose due to the trailing placement-constraint clause, but this does not undermine its distinctiveness or completeness.

Dimension	Reasoning	Score
Specificity	Lists several concrete actions ('Generate new AI images', 'apply those requested changes', 'create revised bitmap images', 'place each result beside the corresponding original or in a nearby clear area') with only minor coverage gaps, though the trailing constraint clause is more of a guardrail than an action.	4 / 5
Completeness	Explicitly answers both 'what' (generate revised AI bitmap images and place them beside originals) and 'when' ('Use when the user provides one or more screenshots showing Cowart images marked with the 批注 tool, arrows, or visible edit notes') with concrete trigger phrases.	5 / 5
Trigger Term Quality	Good natural keyword coverage ('annotation screenshots', '批注 tool', 'arrows', 'visible edit notes', 'screenshots') that a user would plausibly say, though a few common synonyms are absent and the vocabulary is somewhat domain-specialized.	4 / 5
Distinctiveness Conflict Risk	Occupies a clear niche (Cowart 批注 annotation-driven image editing) with distinct triggers and minimal overlap risk against generic image-generation skills.	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: zhongerxin/Cowart
Path: skills/cowart-image-edit/SKILL.md
Commit: c40d254

Reviewed: 7 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.