Optionally checks, then commits code to the current or a new feature branch.
68
52%
Does it follow best practices?
Impact
94%
1.30xAverage score across 3 eval scenarios
Passed
No known issues
Optimize this skill with Tessl
npx tessl skill review --optimize ./.claude/skills/commit/SKILL.mdQuality
Discovery
32%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 conveys the basic idea of committing code with optional checks, but it is too terse and lacks explicit trigger guidance ('Use when...'). It would benefit from more specific action verbs, natural user keywords (like 'git', 'commit message'), and a clear 'when to use' clause to help Claude distinguish it from other version control skills.
Suggestions
Add an explicit 'Use when...' clause, e.g., 'Use when the user wants to commit code, create a feature branch, or save changes to git.'
Include natural trigger terms users would say, such as 'git', 'commit', 'push', 'new branch', 'save changes', 'check in code'.
Clarify what 'checks' means—e.g., 'Runs linting and tests before committing' or 'Validates staged changes'—to improve specificity and distinctiveness.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (git/version control) and some actions (checks, commits, feature branch), but doesn't elaborate on what 'checks' means or list comprehensive actions like staging, diffing, or creating PRs. | 2 / 3 |
Completeness | Describes what it does (checks and commits code to branches) but has no explicit 'Use when...' clause or trigger guidance, which per the rubric should cap completeness at 2, and the 'what' is also somewhat vague, placing this at 1. | 1 / 3 |
Trigger Term Quality | Includes natural terms like 'commits', 'feature branch', and 'code', but misses common variations users might say such as 'git', 'push', 'save changes', 'check in', 'branch off', or 'staged changes'. | 2 / 3 |
Distinctiveness Conflict Risk | Somewhat specific to git commit workflows with the mention of 'feature branch', but could overlap with other git-related skills (e.g., a branching skill, a code review skill, or a general git skill) due to lack of precise scoping. | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
72%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
A well-structured, concise skill that clearly outlines the commit workflow with appropriate branching logic (check vs force, main branch detection). Its main weaknesses are the lack of concrete examples (e.g., a sample commit message, exact check commands) and missing validation after the commit step. The 'check' argument is underspecified—it mentions linting, building, and testing but doesn't indicate what commands to run.
Suggestions
Add a concrete example of a generated commit message to illustrate the expected conventional commit format (e.g., 'feat(auth): add JWT token validation middleware').
Specify what check commands to run for the 'check' argument, or note that they should be discovered from the project's configuration (package.json scripts, Makefile targets, etc.).
Add a brief validation step after committing, such as running 'git log -1' to confirm the commit was created successfully.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every line serves a purpose. No unnecessary explanations of what git is or how commits work. The instructions are lean and assume Claude's competence with git. | 3 / 3 |
Actionability | Provides specific git commands and clear instructions, but lacks executable code examples (e.g., exact git commands with arguments for staging, branch naming conventions, or a concrete commit message example). The conventional commit format is mentioned but no example output is shown. | 2 / 3 |
Workflow Clarity | Steps are clearly sequenced and the check/force argument provides a branching path. However, there's no validation checkpoint after the commit (e.g., verifying the commit succeeded, checking nothing was accidentally staged), and the 'check' argument mentions running checks but doesn't specify what commands to run or what to do on partial failure. | 2 / 3 |
Progressive Disclosure | This is a simple, single-purpose skill under 50 lines. The content is well-organized with clear sections (Arguments, Steps) and doesn't need external references. The structure is appropriate for the complexity. | 3 / 3 |
Total | 10 / 12 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 — 11 / 11 Passed
Validation for skill structure
No warnings or errors.
d93dc41
Table of Contents
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.