Skills diagnose
📦

diagnose

Content revision r1 Safe 🌐 Network access⚙️ External commands

Diagnose Bugs With a Reproducible Loop

Hard bugs waste time when teams guess before reproducing the failure. This skill guides Claude, Codex, or Claude Code from reproduction to regression test.

Supports: Claude Codex Code(CC)
📊 76 Adequate

Install with my Agent

Copy this request to your Agent. It includes the canonical Skill page and manifest.

Agent request
Review the Skillstore skill "diagnose" from https://skillstore.io/skills/mattpocock-diagnose.md and its manifest at https://skillstore.io/api/skills/mattpocock-diagnose/manifest. Verify the artifact. You may proceed after verification, subject to the environment's own policy.

Your Agent should still show its plan and request any confirmation required by the security policy.

Agent-readable resources

Use these links when an AI agent, crawler, or script needs clean context instead of reading the full page.

Markdown detail GET /skills/mattpocock-diagnose.md Signed manifest GET /api/skills/mattpocock-diagnose/manifest Signed lockfile GET /api/skills/mattpocock-diagnose/lockfile

Test it

Using "diagnose". A test fails only on CI after a dependency update.

Expected outcome:

  • A local loop that reproduces the CI-like failure.
  • A ranked hypothesis list tied to dependency behavior.
  • A regression test plan and cleanup checklist.

Using "diagnose". The export button throws an error for some users.

Expected outcome:

  • A browser or human-in-the-loop reproduction plan.
  • A request for the exact symptom and affected conditions.
  • A fix path verified against the original export scenario.

Using "diagnose". A page became slower after a recent release.

Expected outcome:

  • A baseline measurement before any optimization.
  • Focused probes for likely regression causes.
  • A final check that performance improved without new failures.

Security Audit

Safe
v4 • 7/6/2026 Open versioned report

No malicious intent or prompt injection was found in the reviewed files. The static findings are Markdown examples, localhost prompts, or local debugging instructions, not hidden execution or external data transfer. The skill may guide users to run normal project tests or scripts during diagnosis.

2
Files scanned
160
Lines analyzed
0
Review items
0
False positives ignored

Risk Factors

🌐 Network access (1)
scripts/hitl-loop.template.sh:31
⚙️ External commands (6)
SKILL.md:27 SKILL.md:29 SKILL.md:87 SKILL.md:89 SKILL.md:113 SKILL.md:117
No confirmed security findings were detected by the latest completed static and semantic audit. This does not prove the skill has no side effects.
Audited by: codex View Audit History →
Share & cite this report

Share the versioned assessment report, neutral badge, embed card, and citations. Skillstore reports evidence without deciding whether this Skill is safe.

Open versioned report
Security Assessment

Copy report link

https://skillstore.io/skills/mattpocock-diagnose/audits/4?utm_source=security_passport&utm_medium=share&utm_campaign=versioned_report

Markdown badge

[![Skillstore security assessment](https://skillstore.io/badges/skills/mattpocock-diagnose/security.svg)](https://skillstore.io/skills/mattpocock-diagnose?utm_source=security_passport_badge)

HTML badge

<a href="https://skillstore.io/skills/mattpocock-diagnose?utm_source=security_passport_badge"><img src="https://skillstore.io/badges/skills/mattpocock-diagnose/security.svg" alt="Skillstore security assessment" loading="lazy"></a>

Embed card

<iframe src="https://skillstore.io/embed/skills/mattpocock-diagnose.html" title="Skillstore Security Assessment" sandbox="allow-popups allow-popups-to-escape-sandbox" loading="lazy" referrerpolicy="no-referrer" width="420" height="180"></iframe>
Academic citations (APA · BibTeX · CFF)

APA citation

mattpocock. (2026). diagnose security audit report (audit version 4) [Author version unspecified]. Skillstore. https://skillstore.io/skills/mattpocock-diagnose/audits/4

BibTeX citation

@techreport{mattpocock-mattpocock-diagnose-2026, author = {mattpocock}, title = {diagnose security audit report (audit version 4)}, institution = {Skillstore}, year = {2026}, number = {4}, url = {https://skillstore.io/skills/mattpocock-diagnose/audits/4}, note = {Author version unspecified} }

CITATION.cff

cff-version: 1.2.0 message: "If you use this Skill, cite its author and this versioned security audit report." title: "diagnose security audit report (audit version 4)" version: "unspecified" type: report authors: - name: "mattpocock" date-released: "2026-07-06" url: "https://skillstore.io/skills/mattpocock-diagnose/audits/4" identifiers: - type: other value: "skillstore:mattpocock-diagnose:audit:4" description: "Skillstore immutable audit report identifier"

Skillstore Score

Why this score Evidence Confidence: High
45
Architecture
85
Maintainability
87
Content
68
Community
91
Spec Compliance

What You Can Build

Reproduce Application Bugs

Turn vague bug reports into a deterministic loop, then verify the fix against the original symptom.

Isolate Performance Regressions

Measure a baseline, rank likely causes, and test one performance hypothesis at a time.

Structure Debugging Handoffs

Capture the repro, hypotheses, fix evidence, and prevention notes for team review.

Try These Prompts

Start a Basic Diagnosis 
Use the diagnose skill on this bug. Build a feedback loop first. Reproduce the failure before proposing a fix.
Debug a UI Failure 
Diagnose this UI bug. Choose a browser or human-in-the-loop reproduction path. Confirm the visible symptom before changing code.
Investigate a Regression 
Use the diagnose process for this regression. Build a bisectable signal. Rank hypotheses and test one prediction at a time.
Run Full Root Cause Analysis 
Diagnose this failure end to end. Minimize the repro, test hypotheses, fix it, remove instrumentation, and write prevention notes.

Best Practices

  • Start with the fastest deterministic loop that reproduces the real symptom.
  • Write falsifiable hypotheses before changing code or adding instrumentation.
  • Remove temporary logs, prototypes, and debug helpers before declaring the fix complete.

Avoid

  • Changing code before reproducing the original failure.
  • Logging everything instead of adding probes tied to specific hypotheses.
  • Adding shallow tests that do not exercise the real bug path.

Frequently Asked Questions

What problem does this skill solve?
It gives agents a disciplined process for hard bugs and performance regressions. The process starts with a reliable feedback loop.
Does it fix bugs automatically?
No. It guides investigation, hypothesis testing, implementation, and verification using the project context available to the agent.
Can it handle flaky bugs?
Yes. It focuses on raising the reproduction rate until the failure is frequent enough to debug.
Why does it require ranked hypotheses?
Ranked hypotheses reduce guessing. Each hypothesis must predict what evidence would confirm or disprove it.
Does it support performance work?
Yes. It recommends baseline measurement, targeted profiling, and verification before and after the fix.
What access does it need?
It works best with tests, logs, fixtures, traces, or a local environment that reproduces the reported issue.

Developer Details

Author

mattpocock

License

MIT

Skillstore revision

r1

Version notice

The author did not declare a version.

Repository

https://github.com/mattpocock/skills/tree/main/skills/engineering/diagnose/

Ref

b8ca75d2c0a7e7102978993058777d82b8ab2610

Maintenance freshness

7/18/2026

Usage

3 downloads · 80 views

File structure

📁 scripts/

📄 hitl-loop.template.sh

📄 SKILL.md