debug-mode
Skill0 starsChanged 3 months agoMITView on GitHub
Diagnose and fix runtime bugs using temporary structured logs and verified evidence. Use when the user invokes debug-mode or requests instrumented debugging, and resume active sessions after reproduction or verification replies. Use trace-mode for code location without a fix.
8 other projects use a version of this.
Adopted by
Most recently taken up by
- doraemonkeys/claude-code-debug-modetheir own version, first seen Oct 24, 2026
- whoaskedssselery/claude-workspacetheir own version, first seen Oct 20, 2026
- liush2yuxjtu/pi-debug-modetheir own version, first seen Oct 17, 2026
- kaushik912/my-claude-skillstheir own version, first seen Sep 28, 2026
- heiets/universal-debug-mode-skilltheir own version, first seen Sep 14, 2026
- vanenshi/skillstheir own version, first seen Aug 5, 2026
Other versions8 projects use a version of this.
Word for word · 0
None found.
Edited · 0
None found.
Wrote their own version · 8
- doraemonkeys/claude-code-debug-mode★ 120same name
- liush2yuxjtu/pi-debug-mode★ 4same name
- vanenshi/skills★ 2same name
- kaushik912/my-claude-skills★ 1same name
- duniecdawid/ultra-claude★ 0same name
- heiets/universal-debug-mode-skill★ 0same name
- khanashes/claude-skills★ 0same name
- whoaskedssselery/claude-workspace★ 0same name
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 repositoryOne other file it gives its agents.
- trace-modeSkill
The file
We split this skill into the rules it gives an agent
9 rules. 0 are common, 5 are distinctive, and 4 are very specific to this project.
0
Common
Most of the 8 edited versions we read keep these.
5
Distinctive
This author's own choices. Worth a look.
4
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Very specific: line 9
Start for $debug-mode, /debug-mode, debug mode, or an instrumented debugging request.
- Distinctive: line 24
Wait for the user's current-attempt result before advancing either manual gate.
- Distinctive: line 30
Read the entry points, handlers, state transitions, logs, and changes relevant to the symptom.
- Distinctive: line 45
Use enough hypotheses and probes to distinguish the cause, without a fixed quota.
- Distinctive: line 46
Prefer events that separate multiple explanations.
- Very specific: line 51
Read trace-mode and use its caller mode with the Bug Card and Hypothesis Matrix.
- Distinctive: line 55
Keep probes and connection checks active through the fix and verification.
- Very specific: line 59
Use references/checkpoints.md for the attempt boundary, supported interaction path, result validation, and phase mappings.
- Very specific: line 76
Read references/troubleshooting.md only when collection, reproduction, or delivery fails.
“Common” is measured: kept by at least 6 in 10 of the 8 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
Debug Mode
Prove a runtime cause, fix it, verify the original behavior, and clean up.
Start for $debug-mode, /debug-mode, debug mode, or an instrumented debugging
request; resume an active session after replies or compaction. An unrelated task
or a checkpoint answer without an active session does not start instrumentation.
The sibling trace-mode supplies the collector and
instrumentation workflow and requires Node.js 24 or later.
Working Contract
Within host permissions, explicit user scope and workflow choices override skill defaults. Continue authorized work without step-by-step approval. Resolve routine unknowns from available evidence; ask only for missing input that changes the next decision. Keep one compact Bug Card and Hypothesis Matrix in session context, updating them instead of repeating the plan. Read references as their stage applies.
Manual reproduction and verification are the default observations for this workflow. Wait for the user's current-attempt result before advancing either manual gate. Tests or arriving logs cannot supply that result. An explicitly user-selected automated method changes only the named phase; report which method actually ran.
1. Analyze and Hypothesize
Read the entry points, handlers, state transitions, logs, and changes relevant to
the symptom. Keep a Bug Card: symptom versus expected behavior; trigger and
input; candidate path and disputed boundaries; environment and available
evidence; observable acceptance criterion. Mark unavailable facts unknown.
Rank hypotheses supported by that evidence. Each needs:
H1 <cause>
Prediction: <runtime state, path, or timing>
Probe: <events and code locations>
CONFIRMED when: <record condition>
REJECTED when: <record condition>Use enough hypotheses and probes to distinguish the cause, without a fixed quota. Prefer events that separate multiple explanations. Batch independent reads; keep session edits, checkpoints, and cleanup under one owner.
2. Instrument Through Trace Mode
Read trace-mode and use its caller mode with the Bug
Card and Hypothesis Matrix. Read references/probes.md
for the debug event extension. Return before standalone capture with active
probes, the Collector Contract, producer connection status, and any tunnel handle.
Keep probes and connection checks active through the fix and verification.
3. Reproduce, Prove, Fix, and Verify
Use references/checkpoints.md for the attempt
boundary, supported interaction path, result validation, and phase mappings.
PENDING: keep the session active and wait.- Pre-fix
BorC: follow the result table, adjust the attempt, and checkpoint again. - Pre-fix
A: followreferences/prove-fix-verify.mdthrough evidence, correction, verification, and cleanup. No additional approval is needed to fix the confirmed cause within the requested scope.
Completion and Exceptions
Report the cause with current-attempt evidence and file locations, what changed, verification performed, cleanup status, and remaining uncertainty. Use the user's language and detail proportional to the diagnosis.
The session ends after cleanup or explicit abandonment. For abandonment, follow
references/cleanup.md without claiming a verified fix.
Read references/troubleshooting.md only when
collection, reproduction, or delivery fails.
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.

