Diagnosing bugs

A discipline for hard bugs. Skip phases only when explicitly justified. A diagnosis-only request ends with evidence and a proposed fix.

Read Fix Root Causes. For regression tests, follow tdd and Test Behavior, Not Implementation.

For an existing capture, follow trace forensics instead of requiring a new reproduction loop. For live profiling, follow runtime forensics. Those paths may return a provisional diagnosis; they do not establish that a fix works.

The phases below apply to a runnable failure. Reuse reproduction evidence already gathered by the caller. Honor its scope: a read-only inquiry permits existing checks and evidence analysis, not code edits or instrumentation. When a needed experiment exceeds that scope, return the gap and proposed experiment. For a requested fix, bug-fix owns delivery; for optimization, perf-issue owns the measurement comparison. Continue the current diagnosis without recursively restarting those workflows.

For a session transfer, read handoff. Include the reproduction command or capture, hypotheses tested, remaining probes, and cleanup state.

When exploring the codebase, read CONTEXT.md (if it exists) to get a clear mental model of the relevant modules, and check ADRs in the area you're touching.

Redact

This skill has you show commands, outputs and captured artifacts. Redact every secret first: write <REDACTED> in its place. Build loops against env vars, so the credential stays in the environment rather than in what you show. Captured artifacts carry auth headers: quote only the lines that carry the signal.

If the redacted output is not enough to diagnose the bug, say so and ask the user.

Phase 1: Build a feedback loop

This is the skill. Everything else is mechanical. If you have a tight pass/fail signal for the bug (one that goes red on this bug), you will find the cause; bisection, hypothesis-testing, and instrumentation all just consume it. If you don't have one, no amount of staring at code will save you.

Spend disproportionate effort here. Be aggressive. Be creative. Refuse to give up.

Ways to construct one, in roughly this order

1. Failing test at whatever seam reaches the bug: unit, integration, e2e.
2. Curl / HTTP script against a running dev server.
3. CLI invocation with a fixture input, diffing stdout against a known-good snapshot.
4. Headless browser script (Playwright / Puppeteer) that drives the UI and asserts on DOM/console/network.
5. Replay a captured trace. Save a real network request / payload / event log to disk; replay it through the code path in isolation.
6. Throwaway harness. Spin up a minimal subset of the system (one service, mocked deps) that exercises the bug code path with a single function call.
7. Property / fuzz loop. If the bug is "sometimes wrong output", run 1000 random inputs and look for the failure mode.
8. Bisection harness. If the bug appeared between two known states (commit, dataset, version), automate "boot at state X, check, repeat" so you can git bisect run it.
9. Differential loop. Run the same input through old-version vs new-version (or two configs) and diff outputs.
10. HITL bash script. Last resort. If a human must click, drive them with scripts/hitl-loop.template.sh so the loop is still structured. Copy the template and edit its steps for the authorized environment before running it. Capture observations, and leave signing in to the user as a step. Captured output feeds back to you.

Build a loop that reaches the actual failure before choosing a fix.

Tighten the loop

Treat the loop as a product. Once you have a loop, tighten it:

• Can I make it faster? (Cache setup, skip unrelated init, narrow the test scope.)
• Can I make the signal sharper? (Assert on the specific symptom, not "didn't crash".)
• Can I make it more deterministic? (Pin time, seed RNG, isolate filesystem, freeze network.)

A 30-second flaky loop is barely better than no loop; a 2-second deterministic one is tight, a debugging superpower.

Non-deterministic bugs

The goal is not a clean repro but a higher reproduction rate. Loop the trigger 100×, parallelise, add stress, narrow timing windows, inject sleeps. Measure the reproduction rate and record the workload, seed, and trial count. Raise the rate without changing the failure being investigated.

When you genuinely cannot build a loop

Say so explicitly and list what you tried. Ask for the missing environment access or a redacted captured artifact (HAR file, log dump, core dump, or screen recording with timestamps). Continue useful read-only investigation, but label its hypotheses unverified. Production instrumentation requires explicit authorization. Do not treat the reproduction gate as passed without a loop.

Completion criterion: a tight loop that goes red

Phase 1 is done when the loop is tight and red-capable: you can name one command (a script path, a test invocation, a curl) that you have already run at least once (show the invocation and its output, redacted), and that is:

• Red-capable: it drives the actual bug code path and asserts the user's exact symptom, so it can go red on this bug and green once fixed. Not "runs without erroring"; it must be able to catch this specific bug.
• Deterministic: same verdict every run (flaky bugs: a pinned, high reproduction rate, per above).
• Fast: seconds, not minutes.
• Agent-runnable: you can run it unattended; a human in the loop only via scripts/hitl-loop.template.sh.

Read enough code to construct the loop. Treat early theories as provisional until runtime evidence supports them. Proceed to Phase 2 only with the red-capable command; otherwise use the missing-loop procedure above.

Phase 2: Reproduce + minimise

Run the loop. Watch it go red as the bug appears.

Confirm:

• The loop produces the failure mode the user described, not a different failure that happens to be nearby. Wrong bug = wrong fix.
• The failure is reproducible across multiple runs (or, for non-deterministic bugs, reproducible at a high enough rate to debug against).
• You have captured the exact symptom (error message, wrong output, slow timing) so later phases can verify the fix actually addresses it.

Minimise

Once it's red, shrink the repro to the smallest scenario that still goes red. Cut inputs, callers, config, data, and steps one at a time, re-running the loop after each cut, and keep only what's load-bearing for the failure.

Why bother: a minimal repro shrinks the hypothesis space in Phase 3 (fewer moving parts left to suspect) and becomes the clean regression test in Phase 5.

Done when every remaining element is load-bearing: removing any one of them makes the loop go green.

Do not proceed until you have reproduced and minimised.

Phase 3: Hypothesise

Generate competing ranked hypotheses before testing any of them. Aim for 3–5 when the evidence supports distinct explanations; do not invent alternatives to meet a quota. Single-hypothesis generation anchors on the first plausible idea.

Each hypothesis must be falsifiable: state the prediction it makes.

Format: "If <X> is the cause, then <changing Y> will make the bug disappear / <changing Z> will make it worse."

If you cannot state the prediction, the hypothesis is a vibe: discard or sharpen it.

Show the ranked list to the user before testing. They often have domain knowledge that re-ranks instantly ("we just deployed a change to #3"), or know hypotheses they've already ruled out. Cheap checkpoint, big time saver. Don't block on it; proceed with your ranking if the user is AFK.

Phase 4: Instrument

Each probe must map to a specific prediction from Phase 3. Change one variable at a time.

Tool preference:

1. Debugger / REPL inspection if the env supports it. One breakpoint beats ten logs.
2. Targeted logs at the boundaries that distinguish hypotheses.
3. Never "log everything and grep".

Tag every debug log with a unique prefix, e.g. [DEBUG-a4f2]. Cleanup at the end becomes a single grep. Untagged logs survive; tagged logs die.

Perf branch. For performance regressions, logs are usually wrong. Instead: establish a baseline measurement (timing harness, performance.now(), profiler, query plan), then bisect. Measure first, fix second.

Phase 5: Fix + regression test

Continue into implementation only when it is authorized. For diagnosis-only work, report the supported cause, evidence, unresolved hypotheses, and proposed fix or next experiment. Then perform the cleanup below that applies to your probes.

Write the regression test before the fix, but only if there is a correct seam for it.

A correct seam is one where the test exercises the real bug pattern as it occurs at the call site. If the only available seam is too shallow (single-caller test when the bug needs multiple callers, unit test that can't replicate the chain that triggered the bug), a regression test there gives false confidence.

If no correct seam exists, that itself is the finding. Note it. The codebase architecture is preventing the bug from being locked down. Flag this for the next phase.

If a correct seam exists:

1. Turn the minimised repro into a failing test at that seam.
2. Watch it fail.
3. Apply the fix.
4. Watch it pass.
5. Re-run the Phase 1 feedback loop against the original (un-minimised) scenario.

Phase 6: Cleanup

After a fix:

• Original repro no longer reproduces (re-run the Phase 1 loop). For intermittent bugs, compare before/after failure rates and trial counts; zero observed failures alone does not prove impossibility.
• Regression test passes (or absence of seam is documented)
• The hypothesis that turned out correct is stated in the commit or PR message, so the next debugger learns

For diagnosis and fixes alike:

• Temporary instrumentation and runtime patches you introduced are removed, preserving unrelated work. Search for your [DEBUG-...] prefix.
• Retain evidence needed to reproduce or explain the finding. Remove disposable experiments, and report retained artifact paths or any cleanup still pending.