safe-refactor

Restructure code while preserving behavior. Use for extraction, consolidation, ownership moves, or cleanup where verification must bracket structural edits.

Quality

61%

Does it follow best practices?

Run evals on this skill

Adds up to 20 points to the overall score

View guide

Securityby

Passed

No findings from the security scan

Fix and improve this skill with Tessl

tessl review fix ./skills/safe-refactor/SKILL.md

Quality

Content

60%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 body is admirably concise and well-structured for a small instruction-only skill, and verification is correctly foregrounded. Its main gap is actionability: the principles describe what to keep in mind but don't give concrete commands, an explicit step sequence, or an error-recovery loop.

Suggestions

Add a concrete minimal workflow as numbered steps (e.g., 1. capture the current behavior proof — tests/commands to re-run; 2. make one structural edit; 3. re-run the proof; 4. fix until it passes) so Claude has an executable sequence rather than principles.

Give an explicit error-recovery checkpoint (e.g., 'If the proof fails after a structural edit, revert or fix the edit and re-run before moving to the next boundary') to satisfy the validation-loop expectation for behavior-preserving edits.

Name specific verification mechanisms (test suite command, build command, or behavior diff) instead of the abstract phrase 'same proof', so the guidance is copy-paste ready.

Dimension	Reasoning	Score
Conciseness	The body is extremely lean — a one-line framing plus six terse bullet principles and a single closing imperative — with no padding and no explanation of concepts Claude already knows; every token earns its place.	5 / 5
Actionability	Guidance is high-level ('Move one ownership boundary at a time', 'Run same proof after change') with no concrete commands, code, tool invocations, or specific steps to execute, fitting the 'minimal concrete guidance' anchor.	2 / 5
Workflow Clarity	A rough sequence is implied (define boundary → establish verification → edit → re-run proof → stop), and verification is present, but steps are implicit and lack explicit checkpoints or an error-recovery loop, so it sits at the 3 anchor rather than above.	3 / 5
Progressive Disclosure	This is a simple sub-50-line skill with no bundle files, well organized into short labeled principles; per the simple-skill exception, that organization alone merits a 5 for progressive disclosure.	5 / 5
	Total	15 / 20 Passed

Description

62%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 is concise, third-person, and covers both what the skill does and when to use it with several concrete action types. Its main weakness is trigger term quality: the 'Use for...' phrasing leans on internal vocabulary rather than natural user phrases and lacks synonyms.

Suggestions

Add natural-language trigger phrases a user would actually say, e.g. 'Use when the user asks to refactor, extract a function, move code between modules, or consolidate duplicate logic while keeping behavior the same'.

Include common synonyms/extensions such as 'inline', 'rename', 'split module', or 'pull up method' to broaden keyword coverage.

Tighten the 'when' clause to surface the behavior-preservation trigger more plainly, e.g. 'Use when restructuring production code where behavior must stay identical.'

Dimension	Reasoning	Score
Specificity	Names the domain (restructuring code) and several concrete action types — 'extraction, consolidation, ownership moves, or cleanup' — with only minor coverage gaps, fitting between the 3 and 5 anchors, leaning toward 4.	4 / 5
Completeness	It answers both 'what' (restructure code while preserving behavior) and 'when' ('Use for extraction, consolidation, ownership moves, or cleanup where verification must bracket structural edits'), so it clears the cap, though the 'when' could be more explicit and user-natural.	4 / 5
Trigger Term Quality	It includes a 'Use for...' clause listing triggers like 'extraction, consolidation, ownership moves, or cleanup', but these are internal jargon rather than natural phrases a user would say, and synonyms are missing.	3 / 5
Distinctiveness Conflict Risk	The behavior-preservation + verification-bracketing framing carves a fairly distinct niche from a generic 'edit code' skill, with only minor overlap risk against closely related refactoring or testing skills.	4 / 5
	Total	15 / 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: JuliusBrussee/caveman
Path: skills/safe-refactor/SKILL.md
Commit: f5d7294

Reviewed: 1 hour 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.