skills.shby @obra

systematic-debugging

Use when encountering any bug, test failure, or unexpected behavior, before proposing fixes

copies: implexa run skills.sh/systematic-debugging

view source

installs

stars

karma

view original SKILL.md from skills.shclick to expand

Systematic Debugging

Overview

Random fixes waste time and create new bugs. Quick patches mask underlying issues.

Core principle: ALWAYS find root cause before attempting fixes. Symptom fixes are failure.

Violating the letter of this process is violating the spirit of debugging.

The Iron Law

NO FIXES WITHOUT ROOT CAUSE INVESTIGATION FIRST

If you haven't completed Phase 1, you cannot propose fixes.

When to Use
Use for ANY technical issue:

Test failures

Bugs in production

Unexpected behavior

Performance problems

Build failures

Integration issues

Use this ESPECIALLY when:

Under time pressure (emergencies make guessing tempting)

"Just one quick fix" seems obvious

You've already tried multiple fixes

Previous fix didn't work

You don't fully understand the issue

Don't skip when:

Issue seems simple (simple bugs have root causes too)

You're in a hurry (rushing guarantees rework)

Manager wants it fixed NOW (systematic is faster than thrashing)

The Four Phases

You MUST complete each phase before proceeding to the next.

Phase 1: Root Cause Investigation

BEFORE attempting ANY fix:

Read Error Messages Carefully

Don't skip past errors or warnings

They often contain the exact solution

Read stack traces completely

Note line numbers, file paths, error codes

Reproduce Consistently

Can you trigger it reliably?

What are the exact steps?

Does it happen every time?

If not reproducible → gather more data, don't guess

Check Recent Changes

What changed that could cause this?

Git diff, recent commits

New dependencies, config changes

Environmental differences

Gather Evidence in Multi-Component Systems

WHEN system has multiple components (CI → build → signing, API → service → database):

BEFORE proposing fixes, add diagnostic instrumentation:

For EACH component boundary:
  - Log what data enters component
  - Log what data exits component
  - Verify environment/config propagation
  - Check state at each layer

Run once to gather evidence showing WHERE it breaks
THEN analyze evidence to identify failing component
THEN investigate that specific component

Example (multi-layer system):

# Layer 1: Workflow
echo "=== Secrets available in workflow: ==="
echo "IDENTITY: ${IDENTITY:+SET}${IDENTITY:-UNSET}"

# Layer 2: Build script
echo "=== Env vars in build script: ==="
env | grep IDENTITY || echo "IDENTITY not in environment"

# Layer 3: Signing script
echo "=== Keychain state: ==="
security list-keychains
security find-identity -v

# Layer 4: Actual signing
codesign --sign "$IDENTITY" --verbose=4 "$APP"

This reveals: Which layer fails (secrets → workflow ✓, workflow → build ✗)

Trace Data Flow

WHEN error is deep in call stack:

See root-cause-tracing.md in this directory for the complete backward tracing technique.

Quick version:

Where does bad value originate?

What called this with bad value?

Keep tracing up until you find the source

Fix at source, not at symptom

Phase 2: Pattern Analysis

Find the pattern before fixing:

Find Working Examples

Locate similar working code in same codebase

What works that's similar to what's broken?

Compare Against References

If implementing pattern, read reference implementation COMPLETELY

Don't skim - read every line

Understand the pattern fully before applying

Identify Differences

What's different between working and broken?

List every difference, however small

Don't assume "that can't matter"

Understand Dependencies

What other components does this need?

What settings, config, environment?

What assumptions does it make?

Phase 3: Hypothesis and Testing

Scientific method:

Form Single Hypothesis

State clearly: "I think X is the root cause because Y"

Write it down

Be specific, not vague

Test Minimally

Make the SMALLEST possible change to test hypothesis

One variable at a time

Don't fix multiple things at once

Verify Before Continuing

Did it work? Yes → Phase 4

Didn't work? Form NEW hypothesis

DON'T add more fixes on top

When You Don't Know

Say "I don't understand X"

Don't pretend to know

Ask for help

Research more

Phase 4: Implementation

Fix the root cause, not the symptom:

Create Failing Test Case

Simplest possible reproduction

Automated test if possible

One-off test script if no framework

MUST have before fixing

Use the superpowers:test-driven-development skill for writing proper failing tests

Implement Single Fix

Address the root cause identified

ONE change at a time

No "while I'm here" improvements

No bundled refactoring

Verify Fix

Test passes now?

No other tests broken?

Issue actually resolved?

If Fix Doesn't Work

STOP

Count: How many fixes have you tried?

If < 3: Return to Phase 1, re-analyze with new information

If ≥ 3: STOP and question the architecture (step 5 below)

DON'T attempt Fix #4 without architectural discussion

If 3+ Fixes Failed: Question Architecture

Pattern indicating architectural problem:

Each fix reveals new shared state/coupling/problem in different place

Fixes require "massive refactoring" to implement

Each fix creates new symptoms elsewhere

STOP and question fundamentals:

Is this pattern fundamentally sound?

Are we "sticking with it through sheer inertia"?

Should we refactor architecture vs. continue fixing symptoms?

Discuss with your human partner before attempting more fixes

This is NOT a failed hypothesis - this is a wrong architecture.

Red Flags - STOP and Follow Process

If you catch yourself thinking:

"Quick fix for now, investigate later"

"Just try changing X and see if it works"

"Add multiple changes, run tests"

"Skip the test, I'll manually verify"

"It's probably X, let me fix that"

"I don't fully understand but this might work"

"Pattern says X but I'll adapt it differently"

"Here are the main problems: [lists fixes without investigation]"

Proposing solutions before tracing data flow

"One more fix attempt" (when already tried 2+)

Each fix reveals new problem in different place

ALL of these mean: STOP. Return to Phase 1.

If 3+ fixes failed: Question the architecture (see Phase 4.5)

your human partner's Signals You're Doing It Wrong

Watch for these redirections:

"Is that not happening?" - You assumed without verifying

"Will it show us...?" - You should have added evidence gathering

"Stop guessing" - You're proposing fixes without understanding

"Ultrathink this" - Question fundamentals, not just symptoms

"We're stuck?" (frustrated) - Your approach isn't working

When you see these: STOP. Return to Phase 1.

Common Rationalizations

Excuse
Reality

"Issue is simple, don't need process"
Simple issues have root causes too. Process is fast for simple bugs.

"Emergency, no time for process"
Systematic debugging is FASTER than guess-and-check thrashing.

"Just try this first, then investigate"
First fix sets the pattern. Do it right from the start.

"I'll write test after confirming fix works"
Untested fixes don't stick. Test first proves it.

"Multiple fixes at once saves time"
Can't isolate what worked. Causes new bugs.

"Reference too long, I'll adapt the pattern"
Partial understanding guarantees bugs. Read it completely.

"I see the problem, let me fix it"
Seeing symptoms ≠ understanding root cause.

"One more fix attempt" (after 2+ failures)
3+ failures = architectural problem. Question pattern, don't fix again.

Quick Reference

Phase
Key Activities
Success Criteria

1. Root Cause
Read errors, reproduce, check changes, gather evidence
Understand WHAT and WHY

2. Pattern
Find working examples, compare
Identify differences

3. Hypothesis
Form theory, test minimally
Confirmed or new hypothesis

4. Implementation
Create test, fix, verify
Bug resolved, tests pass

When Process Reveals "No Root Cause"

If systematic investigation reveals issue is truly environmental, timing-dependent, or external:

You've completed the process

Document what you investigated

Implement appropriate handling (retry, timeout, error message)

Add monitoring/logging for future investigation

But: 95% of "no root cause" cases are incomplete investigation.

Supporting Techniques

These techniques are part of systematic debugging and available in this directory:

root-cause-tracing.md - Trace bugs backward through call stack to find original trigger

defense-in-depth.md - Add validation at multiple layers after finding root cause

condition-based-waiting.md - Replace arbitrary timeouts with condition polling

Related skills:

superpowers:test-driven-development - For creating failing test case (Phase 4, Step 1)

superpowers:verification-before-completion - Verify fix worked before claiming success

Real-World Impact

From debugging sessions:

Systematic approach: 15-30 minutes to fix

Random fixes approach: 2-3 hours of thrashing

First-time fix rate: 95% vs 40%

New bugs introduced: Near zero vs common

don't have the plugin yet? install it then click "run inline in claude" again.

restructured original into implexa's six-component format, extracted implicit decision logic into explicit decision points, added edge cases (rate limits, auth expiry, network timeouts, intermittent issues), formalized inputs with connection requirements, numbered all procedure steps with explicit inputs/outputs, added outcome signals for verification.

systematic-debugging

intent

when you hit a bug, test failure, or unexpected behavior, your instinct is to try fixes fast. don't. random fixes waste time, create new bugs, and mask the real problem. this skill forces you through four phases of disciplined investigation before you touch code. the core principle: ALWAYS find root cause before attempting fixes. symptom fixes are failure. use this for any technical issue, especially when under time pressure (which is exactly when guessing fails worst).

inputs

access to error messages, logs, and stack traces
ability to reproduce the issue locally or in staging
git history and recent commits
codebase with working examples of similar patterns
(optional) multi-component system visibility (CI logs, service dashboards, database query logs, APM tools)
(optional) external references or documentation for the pattern you're implementing
human partner available for architectural discussion if 3+ fix attempts fail

procedure

phase 1: root cause investigation

step 1. read error messages completely

input: error output, logs, stack traces
action: read every word, not just the headline. note line numbers, file paths, error codes. they contain the exact solution 90% of the time.
output: written summary of what error says (not interpretation, what it literally says)

step 2. reproduce consistently

input: issue description, your local environment
action: follow exact steps to trigger issue. does it happen every time? same conditions? document reproduction steps precisely.
output: reliable reproduction steps or "not reproducible yet" + data gathering list
edge case: if not reproducible, issue may be timing-dependent, environmental, or intermittent. gather more data before moving forward (logs over time, different environments, concurrent requests, etc.).

step 3. check recent changes

input: git log, recent commits, dependency changes, config changes
action: what changed that could cause this? git diff against known-good state. compare environments (staging vs prod, old vs new infra).
output: list of recent changes that could affect this behavior

step 4a. gather evidence in single-component systems

input: error output, system logs
action: trace the error backward through the code path. where does the bad value originate? what called this function with bad input? keep tracing up until you hit the entry point.
output: call chain showing where bad data entered the system

step 4b. gather evidence in multi-component systems

input: multi-component system (CI → build → signing, API → service → database, etc.)
action: add diagnostic instrumentation at each component boundary. for each boundary, log what data enters, what data exits, verify environment/config propagation, check state at each layer. run once to gather evidence showing WHERE it breaks, then analyze to identify failing component, then investigate that specific component deeply.
output: evidence log showing which component boundary fails (example: secrets available in workflow ✓, secrets not in build script ✗)
edge case: network timeouts, rate limits on external APIs, auth token expiry between layers, missing env var propagation

step 5. trace data flow (for deep call stack errors)

input: error location in call stack, function signatures
action: follow the root-cause-tracing.md technique. quick version: where does bad value come from? what called this with bad value? keep tracing backward until you find the source. fix at source, not at symptom.
output: identified source of bad value, not the symptom location

phase 2: pattern analysis

step 6. find working examples

input: codebase, similar working implementations
action: locate similar working code in same codebase. what works that's similar to what's broken?
output: working reference implementation identified

step 7. read reference implementation completely

input: reference code
action: read every line. don't skim. understand the pattern fully before applying.
output: written explanation of how reference pattern works and its assumptions

step 8. identify differences

input: broken code, working reference
action: list every difference, however small. don't assume "that can't matter".
output: exhaustive diff list between broken and working

step 9. understand dependencies

input: reference pattern, your system
action: what other components does this pattern need? what settings, config, environment? what assumptions does it make?
output: dependency map and assumptions list

phase 3: hypothesis and testing

step 10. form single hypothesis

input: evidence from phases 1-2, pattern analysis
action: state clearly in writing: "I think X is the root cause because Y". be specific, not vague. this is your testable claim.
output: written hypothesis

step 11. test minimally

input: hypothesis, ability to modify code/config
action: make the SMALLEST possible change to test hypothesis. one variable at a time. don't fix multiple things at once.
output: test change applied, test result (pass or fail)

step 12. verify before continuing

input: test result
action: did it work? yes → go to phase 4. no → go back to step 10 (form NEW hypothesis, don't add more fixes on top).
output: confirmation of hypothesis (pass) or rejection (fail)
edge case: hypothesis failed. count how many fixes you've tried. if < 3, return to phase 1 with new information. if ≥ 3, stop and question architecture (see phase 4.5 below).

phase 4: implementation

step 13. create failing test case

input: reproduction steps from step 2
action: write simplest possible test that reproduces issue. automated test if framework exists, one-off script if not. MUST have failing test before fixing.
output: failing test, verified to fail consistently
external connection: if using test-driven-development skill, follow that process

step 14. implement single fix

input: root cause identified, failing test from step 13
action: address the root cause, not the symptom. ONE change at a time. no "while I'm here" improvements. no bundled refactoring.
output: code change applied

step 15. verify fix

input: test suite, modified code
action: does test pass now? no other tests broken? does issue actually resolve in reproduction steps? if all yes, you're done. if any no, stop and go to step 16.
output: passing test suite, verified issue resolved

step 16. if fix doesn't work

input: failed fix attempt, count of attempts
action: STOP. count how many fixes you've tried. if < 3, return to phase 1 and re-analyze with new information. if ≥ 3, go to step 17.
output: decision to retry phase 1 or question architecture

step 17. if 3+ fixes failed, question architecture

input: multiple failed fix attempts, pattern of failures
action: pattern indicates architectural problem if: each fix reveals new shared state/coupling/problem elsewhere, fixes require massive refactoring, each fix creates new symptoms elsewhere. stop and question fundamentals: is this pattern sound? are we sticking with it through sheer inertia? should we refactor architecture vs. continue fixing symptoms? discuss with human partner before attempting more fixes.
output: architectural discussion, decision to refactor or pivot approach

decision points

if not reproducible (step 2): gather more data (enable debug logging, run in different environments, inject traffic patterns) before moving to phase 2. don't investigate patterns of something you can't trigger.

if multi-component system (step 4b vs 4a): use step 4b. single-component debugging can't isolate failures across layers. multi-component systems REQUIRE instrumentation at each boundary.

if hypothesis failed (step 12): don't add more fixes. form a new hypothesis and test again. each failed test teaches you something.

if fewer than 3 fixes failed (step 16): return to phase 1, not forward to architecture. new evidence from the failed fix may reveal what you missed.

if 3 or more fixes failed (step 16-17): don't attempt fix #4. question architecture first. 3+ failures in different places means the pattern or design is broken, not the implementation.

if issue is truly environmental/timing-dependent (phase 1 conclusion): you've completed the process correctly. document what you investigated. implement appropriate handling (retry logic, timeout handling, better error messages). add monitoring and logging for future investigation. but: 95% of "no root cause" cases are incomplete investigation.

if human partner says "is that not happening?" or "stop guessing": you're proposing fixes without understanding. return to phase 1.

if you're thinking "quick fix for now, investigate later": stop. first fix sets the pattern. do systematic debugging from the start.

if you're thinking "just try changing X and see if it works": stop. this is guess-and-check. go to phase 3, form a hypothesis first.

output contract

success is NOT "bug fixed". success is:

root cause identified and documented: written statement of why the bug happens, traced to its origin (phase 1 complete)
pattern understood: written comparison showing what's different between broken and working (phase 2 complete)
hypothesis tested: single change tested and verified to resolve issue (phase 3 complete)
failing test exists: automated or manual test that reproduces issue and passes after fix (step 13)
fix applied and verified: code changed, test passes, no new failures, issue resolved (step 15)
all tests pass: original test suite runs without regression

file locations:

investigation notes: keep in issue tracker or PR description
failing test: committed to repo in test suite or as standalone script
fix: committed with clear commit message referencing root cause (not symptom)
if architectural refactor decided: separate issue/PR with architectural discussion included

outcome signal

you know it worked when:

you can explain the root cause in one sentence (if you can't, phase 1 incomplete)
you can point to the exact line where bad data originated (not where it broke, where it started)
your test that was failing now passes
you tried ONE fix and it worked (if you tried multiple, process wasn't systematic)
no other tests broke
the actual issue is resolved (manual verification matches test verification)
if 3+ fixes failed, you stopped and questioned architecture instead of trying fix #4

opposite signal: you're done when you realize you're thinking "just one more fix" or "let me try this and see", or when your human partner says "stop guessing" or "we're stuck". these mean return to phase 1.