spectra-debug
Skill44 starsChanged 8 months agoMITView on GitHub
Systematically debug a problem using a four-phase workflow
6 other projects use a version of this, 6 of them edited.
Adopted by
Most recently taken up by
- Will-SW74/passdown-osedited, first seen Oct 11, 2026
- DanSnow/hoarder-pipetteedited, first seen Oct 8, 2026
- leoho0722/naver-blog-image-downloaderedited, first seen Aug 31, 2026
- PsychQuant/claude-hot-limitedited, first seen Jul 1, 2026
- harrychuang/cursor-skillsedited, first seen Jun 17, 2026
- BananaBay69/claude-code-watchdogedited, first seen Apr 29, 2026
Used with
Projects with it also use spectra-commit, spectra-analyze and 2 more.
Other files people use alongside this one.
Files found in the same public projects as a version of this one, more often than they turn up anywhere else. Each project counts once, large collections count for less, and this file's own repository is left out.
BananaBay69Skillspectra-commit
Commit files related to a specific Spectra change
Also in 6 projects that use this
harrychuangSkillspectra-analyze
Analyze artifact consistency for a change
Also in 5 projects that use this
harrychuangSkillspectra-drift
Detect drift between a Spectra change and the current codebase state
Also in 5 projects that use this
- +1 more from harrychuang
Other versions6 projects use a version of this.
Word for word · 0
None found.
Edited · 6
- DanSnow/hoarder-pipette★ 9829% same
- harrychuang/cursor-skills★ 1999% same
- BananaBay69/claude-code-watchdog★ 299% same
- leoho0722/naver-blog-image-downloader★ 299% same
- PsychQuant/claude-hot-limit★ 099% same
- Will-SW74/passdown-os★ 099% same
Wrote their own version · 0
None found.
How matches are found. Word for word: the exact same text. Edited: shares passages with this file (“99% same” means 99% of its five-word phrases also appear here; 10% or more counts). Their own version: similar meaning, or, weakest, the same name. We can’t tell which project had it first. Projects are counted among the 200 most-starred candidates, and mentions are links to the repository we found, up to 50 per site, so both counts are a floor.
In this repository18 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- plugin/CLAUDE.mdCLAUDE.md
- spectra-applySkill
- spectra-archiveSkill
- spectra-askSkill
- spectra-auditSkill
- spectra-debugSkill
- spectra-discussSkill
- spectra-ingestSkill
- spectra-proposeSkill
- spectra-applySkill
- spectra-archiveSkill
- spectra-askSkill
- spectra-auditSkill
- spectra-discussSkill
- spectra-ingestSkill
- spectra-proposeSkill
The file
We split this skill into the rules it gives an agent
7 rules. 7 are common, 0 are distinctive, and 0 are very specific to this project.
7
Common
Most of the 6 edited versions we read keep these.
0
Distinctive
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Common: line 15
This skill enforces debugging discipline: No guessing, no random changes, no "let me try this." Every step is deliberate and evidence-based.
- Common: line 30
Document what you tried and why it failed
- Common: line 34
Do NOT keep trying variations of the same approach.
- Common: line 40
Before anything else, make the bug happen reliably.
- Common: line 47
If you can't reproduce it, you can't debug it.
- Common: line 58
Use git bisect: when the bug is a regression — find the exact commit that introduced it
- Common: line 75
Don't stop at the first explanation: Verify your hypothesis:
“Common” is measured: kept by at least 6 in 10 of the 6 edited versions of this file we could read. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Systematically debug a problem using a four-phase workflow.
This skill enforces debugging discipline. No guessing, no random changes, no "let me try this." Every step is deliberate and evidence-based.
Input: The argument after /spectra-debug describes the bug or unexpected behavior. Examples:
/spectra-debug the search returns duplicate results/spectra-debug crash on startup after upgrading/spectra-debug file watcher misses rename events
The Three-Attempt Rule
Maximum 3 fix attempts per hypothesis in Phase 4 (Fix). Phases 1-3 (Reproduce, Isolate, Root Cause) are investigation — they do not count toward this limit. If your third fix attempt fails:
- Stop fixing
- Document what you tried and why it failed
- Question your hypothesis — is the root cause what you think it is?
- Research alternatives or try a completely different angle
Do NOT keep trying variations of the same approach. That's a loop, not debugging.
Phase 1: Reproduce
Before anything else, make the bug happen reliably.
- Find the exact steps to trigger the bug
- Identify the expected vs actual behavior — be precise
- Determine if it's consistent — does it happen every time? Only on certain input?
- Simplify the reproduction — strip away everything that's not essential
If you can't reproduce it, you can't debug it. Gather more information before proceeding.
Phase 2: Isolate
Narrow down where the bug lives.
- Binary search the codebase — which module, which function, which line?
- Check inputs and outputs — at each boundary, is the data correct?
- Add targeted logging — not everywhere, just at decision points
- Use git bisect when the bug is a regression — find the exact commit that introduced it
Goal: pinpoint the exact location where behavior diverges from expectation.
Phase 3: Root Cause
Understand WHY it's broken, not just WHERE.
Ask these questions:
- What assumption is being violated?
- What changed that made this start failing?
- Is this a symptom of a deeper issue, or the actual problem?
- Are there other places with the same pattern that might also be affected?
Don't stop at the first explanation. Verify your hypothesis:
- Can you predict the bug's behavior based on your theory?
- Does your theory explain ALL the symptoms, not just some?
- Can you construct a test case that proves the root cause?
Phase 4: Fix
Now — and only now — fix the bug.
- Write a failing test that reproduces the bug. If
tdd: trueis set in.spectra.yaml, fetch TDD instructions viaspectra instructions --skill tddand follow the Red-Green-Refactor cycle - Make the minimum change to fix the root cause — not the symptoms
- Run the test — confirm it passes
- Run the full test suite — ensure no regressions
- Check related code — if this pattern exists elsewhere, fix those too
Rationalization Table
| What You're Thinking | What You Should Do |
| "I bet it's this, let me just change it" | Reproduce first. Verify your hypothesis |
| "Let me add some prints everywhere" | Add targeted logging at specific boundaries |
| "It works on my machine" | Find what's different in the failing environment |
| "Let me try reverting this change" | Use git bisect to find the actual cause |
| "The fix is obvious, I don't need a test" | The fix is wrong. Write the test |
| "Let me just restart the service" | That hides the bug. Find the root cause |
| "Maybe if I just clear the cache..." | Understand why the cache was wrong |
Guardrails
- Don't guess — Every change must be based on evidence
- Don't fix symptoms — Find and fix the root cause
- Don't skip the test — Phase 4 always starts with a failing test
- Don't power through — After 3 failed attempts, stop and reassess
- Do keep notes — Document what you tried, what you found, what you ruled out
- Do check broadly — A bug in one place often means the same bug exists elsewhere
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

