Skip to content
Get packs

Session and Repo Debugger

Diagnose stuck or broken Claude Code/OMC sessions using logs, traces, state and focused reproduction

Source Download skill
Repository
Yeachan-Heo/oh-my-claudecode
Stars
40,512
Licence
MIT
Files
1 file

What it does

When installed, the agent approaches a confusing failure in a Claude Code or oh-my-claudecode session in a structured way: it reads the issue, checks the most relevant local evidence such as trace and state data, notes and failing tests, tries to reproduce the problem narrowly, and separates the visible symptom from the likely root cause. It then recommends the smallest next fix or check, and it tags any temporary debug output so it can be cleaned up easily.

Great for

  • Developers whose Claude Code or OMC workflow, hooks or multi-agent orchestration is behaving unexpectedly
  • Maintainers who want a consistent debugging checklist with evidence before fixes
  • Teams that need a reminder to redact secrets from logs before sharing them

Skip it if

  • Everyday app bugs unrelated to agent tooling: the skill is aimed at session and runtime problems, so general coding help may be better served by other tools
  • Non-technical users who have no access to logs or a local repo, since the workflow depends on inspecting local evidence
  • Situations where logs contain secrets and the agent is asked to share them externally without review; the skill asks for redaction but the user should still check output

Try asking

  • “My OMC workflow stopped delegating to agents after the last update. Help me find the root cause.”
  • “Hooks seem to fire twice in my Claude Code session. Trace what's happening and tell me the smallest fix.”
  • “Tests pass locally but the agent flow fails in orchestration. Diagnose the failure signal.”
  • “Something in my project memory looks wrong after a session. Inspect the state and explain what broke.”
Step-by-step guide: how to use Session and Repo DebuggerInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 1 original files, the MIT licence and an attribution note.

    debug.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip debug.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip debug.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 3

    Ask for the task

    No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.

Get better results

  • Describe the exact symptom, when it started, and the command or workflow involved so the agent can narrow the reproduction.
  • Ask the agent to show its evidence, since the output format asks for observed failure, hypothesis, supporting evidence and next action.
  • Review any debug output before sharing it, even though the skill instructs redaction of secrets.

What's inside

  • SKILL.md

Redistributed unmodified from Yeachan-Heo/oh-my-claudecode under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.