Diagnoses bugs and performance regressions through a structured reproduce-hypothesise-fix workflow.
npx skills add https://github.com/jecg2804/HumanOS --skill diagnose-jecg2804
Please help me install this Agent Skill. Skill: diagnose Source: https://github.com/jecg2804/HumanOS/tree/main/.claude/skills/diagnose Command: npx skills add https://github.com/jecg2804/HumanOS --skill diagnose-jecg2804
[DEBUG-xxxx] prefix so cleanup is a single grep, and the post-mortem phase records the confirmed root cause in the commit message. - Use Case: A user reports that an export button intermittently throws an error. The Skill builds a reproduction loop (using the included human-in-the-loop bash template if needed), reproduces the failure, tests ranked hypotheses one variable at a time, and lands a fix guarded by a regression test. ## Quick Start Ask the agent to diagnose the bug you are experiencing and describe the exact symptom, error message, or performance regression you observed.High-intent search queries and answers about installing and using this skill.
Raise the reproduction rate instead of chasing a clean repro: loop the trigger 100 times, parallelise, add stress, narrow timing windows, or inject sleeps. A bug that fails 50% of the time is debuggable; one that fails 1% of the time is not, so keep increasing the rate until hypotheses can be tested against it.
Establish a baseline measurement first using a timing harness, performance.now(), a profiler, or a query plan, then bisect to find the cause. Logs are usually the wrong tool for performance work; measure first and fix second, changing one variable at a time.
Stop and say so explicitly rather than guessing. List what you tried, then ask for environment access, a captured artifact such as a HAR file, log dump, or screen recording, or permission to add temporary production instrumentation before forming hypotheses.
Write the regression test before the fix, but only when a correct seam exists where the test exercises the real bug pattern as it occurs at the call site. If no correct seam exists, that absence is itself a finding about the codebase architecture and should be flagged.
Generating 3–5 ranked falsifiable hypotheses prevents anchoring on the first plausible idea. Each hypothesis must state a testable prediction, and showing the ranked list to the user lets domain knowledge re-rank it cheaply before any testing time is spent.
anderson-joyleReproduce software bugs and generate ranked falsifiable hypotheses.
janpeterdLocate shallow modules and propose deepened interfaces for codebase architecture.
robkamEnforce a red-green-proof loop starting with a failing regression test.
elizaOSAutomate cargo-fuzz fuzzing campaigns with libFuzzer and sanitizer instrumentation.