diagnose
Disciplined diagnosis loop for hard bugs and performance regressions. Reproduce → minimise → hypothesise → instrument → fix → regression-test. Use when user says "diagnose this" / "debug this", reports a bug, says something is broken/throwing/failing, or describes a performance regression.
77
96%
Does it follow best practices?
Impact
—
No eval scenarios have been run
Passed
No known issues
Quality
Discovery
100%Based on the skill's description, can an agent find and select it at the right time? Clear, specific descriptions lead to better discovery.
This is a strong skill description that clearly communicates a structured debugging methodology with specific steps, and provides explicit trigger guidance covering multiple natural phrasings users would employ. It uses proper third-person voice and is concise without being vague. The only minor note is that it could potentially overlap with a general 'fix code' skill, but the emphasis on the disciplined diagnosis loop and performance regressions provides sufficient distinction.
| Dimension | Reasoning | Score |
|---|---|---|
Specificity | Lists a concrete, multi-step methodology: 'Reproduce → minimise → hypothesise → instrument → fix → regression-test.' These are specific, actionable steps that clearly describe what the skill does. | 3 / 3 |
Completeness | Clearly answers both 'what' (disciplined diagnosis loop with explicit steps) and 'when' (explicit 'Use when' clause listing multiple trigger scenarios including bug reports, broken/throwing/failing states, and performance regressions). | 3 / 3 |
Trigger Term Quality | Includes a strong set of natural trigger terms users would actually say: 'diagnose this', 'debug this', 'bug', 'broken', 'throwing', 'failing', 'performance regression'. These cover common variations of how users describe debugging needs. | 3 / 3 |
Distinctiveness Conflict Risk | Targets a clear niche — hard bugs and performance regressions with a structured diagnosis methodology. The specific trigger terms ('diagnose', 'debug', 'broken', 'performance regression') distinguish it well from general coding or testing skills. | 3 / 3 |
Total | 12 / 12 Passed |
Implementation
92%Reviews the quality of instructions and guidance provided to agents. Good implementation is clear, handles edge cases, and produces reliable results.
This is an excellent methodology skill that provides a rigorous, phased debugging discipline with clear gates, validation checkpoints, and actionable techniques at every step. The content is lean and respects Claude's intelligence while teaching a specific workflow Claude wouldn't naturally follow. The only weakness is that the content is entirely self-contained with no bundle files to support referenced artifacts (like the HITL template script) or to offload detailed reference material.
Suggestions
Provide the referenced `scripts/hitl-loop.template.sh` as a bundle file, or remove the reference if it's expected to exist in the user's project.
Consider extracting the 10 feedback loop construction techniques into a separate reference file (e.g., FEEDBACK_LOOPS.md) to keep the main skill leaner and improve progressive disclosure.
| Dimension | Reasoning | Score |
|---|---|---|
Conciseness | Every section earns its place. The content teaches a specific debugging discipline that Claude wouldn't inherently follow — the phased approach, the emphasis on feedback loops, the ranked hypothesis generation, the tagged debug logs. No wasted tokens explaining what debugging is or how programming works. | 3 / 3 |
Actionability | Highly actionable throughout: Phase 1 provides 10 concrete, ordered techniques for building feedback loops; Phase 3 gives a specific format for falsifiable hypotheses; Phase 4 specifies tagged log prefixes with grep-based cleanup; Phase 5 gives a precise 5-step sequence. The guidance is specific and directly executable even without code snippets, which is appropriate for a methodology skill. | 3 / 3 |
Workflow Clarity | Six clearly sequenced phases with explicit gate conditions ('Do not proceed until you reproduce the bug'), validation checkpoints (checklists in Phase 2 and Phase 6), and feedback loops (iterate on the loop itself, re-run Phase 1 loop after fix). The error recovery path for non-deterministic bugs and the 'when you genuinely cannot build a loop' fallback are excellent. | 3 / 3 |
Progressive Disclosure | The content is well-structured with clear headers and phases, but it's a single monolithic file with no references to supporting materials. The HITL bash script template (`scripts/hitl-loop.template.sh`) is referenced but not provided in any bundle. The cross-reference to `/improve-codebase-architecture` is good but the skill could benefit from splitting detailed technique lists (e.g., the 10 feedback loop methods) into a reference file. | 2 / 3 |
Total | 11 / 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.
- Repository
- mattpocock/skills
- Commit
0288510
- 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.