autofix

Review and repair current local changes until they converge, or run Qwen Code Autofix issue and review workflows from GitHub Actions.

Quality

69%

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 ./.qwen/skills/autofix/SKILL.md

Quality

Content

77%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.

The content is highly actionable with concrete commands, JSON schemas, and rigorous validation-checkpointed workflows, but it is verbose and monolithic — a large single file where per-mode detail could be split into referenced files for better progressive disclosure.

Suggestions

Split each mode (local working tree, assess-candidates, develop-issue, address-review) into a one-level-deep reference file and keep SKILL.md as a concise overview with clear links, to improve progressive disclosure for a skill this large.

Factor the repeated trusted-verification command list (build/typecheck/lint/Vitest/integration/generate:settings-schema) into a single shared block referenced by each mode instead of restating it verbatim three times.

Dimension	Reasoning	Score
Conciseness	The body is dense and almost entirely load-bearing policy (no concept explanations Claude already knows), but at ~560 lines it repeats the trusted-verification command list verbatim across the GitHub Actions Rules, develop-issue, and address-review sections, and several winding compound sentences could be tightened.	3 / 5
Actionability	Provides fully executable, copy-paste-ready guidance: the exact `qwen review run` command with flags, concrete `npm run build/typecheck/lint` commands, and explicit JSON schemas and file paths for every required output.	5 / 5
Workflow Clarity	The local mode is an 8-step numbered workflow with explicit fail-closed validation checkpoints, fingerprint re-computation, and convergence criteria, and every GitHub Actions mode includes validate-then-fix-then-rerun feedback loops matching the anchor-5 pattern.	5 / 5
Progressive Disclosure	The file is well-organized into clear mode sections, but it is a single monolithic ~560-line SKILL.md with no one-level-deep references to split the per-mode detail into separate files; the bundled scripts/run-agent.mjs is not navigated to from the body.	3 / 5
	Total	16 / 20 Passed

Description

61%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.

The description clearly states what the skill does across its two lanes but omits an explicit "Use when..." trigger clause, which caps completeness. Trigger terms and distinctiveness are reasonably strong.

Suggestions

Add an explicit "Use when..." clause naming the natural trigger phrases (e.g., direct /autofix invocation, or when GitHub Actions runs an Autofix issue/review workflow) to lift completeness above 3.

Name the assess-candidates and address-review sub-modes so the action coverage reads as comprehensive rather than partial.

Dimension	Reasoning	Score
Specificity	Names the domain and a couple of concrete actions ("Review and repair current local changes until they converge", "run Qwen Code Autofix issue and review workflows from GitHub Actions"), but the actions are stated at a high level and coverage is incomplete (no mention of assess-candidates or address-review modes).	3 / 5
Completeness	The "what" is clear, but there is no "Use when..." clause or equivalent explicit trigger guidance, which caps completeness at 3 per the rubric guidelines; "when" is only weakly implied.	3 / 5
Trigger Term Quality	Includes natural terms a user would say ("autofix", "review", "local changes", "GitHub Actions", "issue", "review workflows") with good coverage; a few natural synonyms are missing but the core triggers are present.	4 / 5
Distinctiveness Conflict Risk	The combination of "Qwen Code Autofix", "local changes", and "GitHub Actions issue and review workflows" carves a fairly distinct niche with minimal overlap risk against general review skills.	4 / 5
	Total	14 / 20 Passed

Validation

87%

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 — 14 / 16 Passed

Validation for skill structure

Criteria	Description	Result
skill_md_line_count	SKILL.md is long (566 lines); consider splitting into references/ and linking	Warning
referenced_paths_exist	Referenced path issues: 1 missing	Warning

	Total	14 / 16 Passed

Repository: QwenLM/qwen-code
Path: .qwen/skills/autofix/SKILL.md
Commit: 39a84c9

Reviewed: 4 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.