review-local-changes
Comprehensive review of local uncommitted changes using specialized agents with code improvement suggestions
54
43%
Does it follow best practices?
Impact
Pending
No eval scenarios have been run
Risky
Do not use without reviewing
Optimize this skill with Tessl
npx tessl skill review --optimize ./plugins/review/skills/review-local-changes/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 identifies a reasonably specific domain (reviewing uncommitted code changes) but suffers from vague language ('comprehensive review', 'specialized agents') and completely lacks explicit trigger guidance. It would benefit from listing concrete actions and adding a 'Use when...' clause with natural user terms.
Suggestions
Add a 'Use when...' clause with natural trigger terms like 'review my changes', 'code review', 'git diff', 'what did I change', 'check my code before committing'.
Replace vague phrases like 'comprehensive review' and 'specialized agents' with specific concrete actions such as 'analyzes git diffs, identifies bugs, suggests refactors, checks style issues'.
Include common file/tool references users might mention, such as 'git', 'staged changes', 'uncommitted work', 'pre-commit review'.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Names the domain (local uncommitted changes) and some actions (review, code improvement suggestions), but lacks specific concrete actions like 'analyze diffs', 'check for bugs', 'suggest refactors', etc. 'Comprehensive review' and 'specialized agents' are somewhat vague. | 2 / 3 |
Completeness | Describes what it does (review uncommitted changes with suggestions) but completely lacks a 'Use when...' clause or any explicit trigger guidance for when Claude should select this skill. Per rubric guidelines, missing 'Use when' caps completeness at 2, and the 'what' is also somewhat vague, so this scores a 1. | 1 / 3 |
Trigger Term Quality | Includes some relevant terms like 'uncommitted changes' and 'code improvement suggestions', but misses common natural user phrases like 'review my changes', 'code review', 'git diff', 'staged changes', 'what did I change'. 'Specialized agents' is jargon unlikely to be used by users. | 2 / 3 |
Distinctiveness Conflict Risk | The focus on 'local uncommitted changes' provides some distinctiveness, but 'code improvement suggestions' is broad enough to overlap with general code review or linting skills. The mention of 'specialized agents' adds some uniqueness but is unclear. | 2 / 3 |
Total | 7 / 12 Passed |
Implementation
55%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This skill is highly actionable with an excellent multi-phase workflow and clear validation/filtering logic, but it suffers significantly from verbosity and poor progressive disclosure. The content is roughly 3-4x longer than necessary, with redundant explanations of filtering logic, inline output templates, and concepts Claude already understands. Splitting templates and detailed rubrics into separate files would dramatically improve token efficiency.
Suggestions
Extract the JSON and markdown output templates into separate referenced files (e.g., TEMPLATES.md) to reduce the main skill's token footprint by ~30%.
Remove redundant explanations of MIN_IMPACT_SCORE resolution — it's defined in the Configuration Resolution block, then re-explained in Phase 3, then demonstrated with a concrete example. One definition plus one example is sufficient.
Cut the false positives section and evaluation guidelines significantly — most items (e.g., 'pre-existing issues in unchanged code', 'pedantic nitpicks') are things Claude already understands as a code reviewer. Reduce to a concise bullet list of the non-obvious ones only.
Move the confidence/impact scoring rubrics and the progressive threshold table into a separate SCORING.md file, keeping only a brief summary in the main skill.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | The skill is extremely verbose at ~300+ lines. It over-explains filtering logic with redundant concrete examples, restates concepts multiple times (e.g., MIN_IMPACT_SCORE resolution is explained in three separate places), and includes lengthy tables and rubrics that Claude could infer from simpler descriptions. The false positives section explains things Claude already knows. | 1 / 3 |
Actionability | The skill provides highly concrete, executable guidance: specific git commands, exact agent prompts to use, precise scoring rubrics with numeric thresholds, complete JSON and markdown output templates, and detailed filtering logic with worked examples. Every step is copy-paste actionable. | 3 / 3 |
Workflow Clarity | The three-phase workflow (Preparation → Issue Search → Scoring/Filtering) is clearly sequenced with explicit validation checkpoints. The progressive confidence threshold table provides a clear feedback mechanism, filter application order is explicitly stated with a concrete example, and the phase transitions are well-defined. | 3 / 3 |
Progressive Disclosure | The entire skill is a monolithic wall of text with no references to external files. The JSON/markdown templates, scoring rubrics, false positive examples, and evaluation guidelines are all inline. The argument parsing section, output templates, and agent instructions could easily be split into separate referenced files to reduce cognitive load. | 1 / 3 |
Total | 8 / 12 Passed |
Validation
90%Checks the skill against the spec for correct structure and formatting. All validation checks must pass before discovery and implementation can be scored.
Validation — 10 / 11 Passed
Validation for skill structure
| Criteria | Description | Result |
|---|---|---|
frontmatter_unknown_keys | Unknown frontmatter key(s) found; consider removing or moving to metadata | Warning |
Total | 10 / 11 Passed | |
- Repository
- NeoLabHQ/context-engineering-kit
- Commit
dedca19
- 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.