DebugCAUTION
The Claude Code background agent for Linear, Slack, Github, GitLab etc. you deploy anywhere. Supports Codex, Cursor, Gemini, and Opencode harnesses too.
Overview
The Claude Code background agent for Linear, Slack, Github, GitLab etc. you deploy anywhere. Supports Codex, Cursor, Gemini, and Opencode harnesses too.
a792ab0e17bbOBSERVED · 2026-10-13What it tells the agent
The instruction file, verbatim from the audited commit — this is the text the model reads, and the surface the audit's instruction layer examines. Quoted here so you can judge it without cloning anything.
--- name: debug description: Full debugging workflow — reproduce the bug with a focused check, perform root cause analysis, then implement a minimal fix. --- # Debug Debug a reported issue using a structured reproduction-then-fix approach. ## Phase 1: Reproduction & Root Cause Analysis 1. **Investigate** — Analyze the bug report for key symptoms and error messages 2. **Trace** — Search the codebase for error occurrence patterns, trace from symptom to source 3. **Root cause** — Identify the root cause through data flow analysis and edge case checking 4. **Reproduce** — Reproduce the exact error with the smallest useful check (a command, dry run, manual flow, or automated test) 5. **Document** — Clearly document the root cause and reproduction steps ## Phase 2: Fix Implementation 1. **Plan** — Analyze the optimal fix approach based on root cause, check for similar fixes in the codebase 2. **Implement** — Make the minimal, targeted fix that addresses the root cause 3. **Verify** — Repeat the reproduction to confirm the fix and run relevant existing checks. Follow `verify-and-ship` for proportionate scope; add a regression test only when the concrete failure warrants durable coverage ## Principles - **Minimal changes** — Fix the bug, nothing more - **Targeted** — Only touch affected code paths - **Verified** — Confirm the reproduction succeeds after the fix and preserve relevant existing safeguards; a new test file is not mandatory - **No unrelated improvements** — Stay focused on the specific bug
Trust audit
CAUTIONgrade B · trust 89/100 Install with care. The audit found things worth knowing before you trust its output.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | WARN |
| L1 | Static analysis of the code | NA |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- none-observed
- Shell
- none-observed
- Dependencies
- pinned
- Secrets in source
- none-found
Findings (5)
.claude/skills/cyrus-setup
.claude/skills/cyrus-setup-claude-auth
.claude/skills/cyrus-setup-endpoint
.claude/skills/cyrus-setup-github
.claude/skills/cyrus-setup-gitlab
Gates applied: no_behavioural_pass.
a792ab0e17bbfull audit observations/trust-audit/skill/cyrusagents__debug.json · Report an issue / request a re-scanAudit history
Every audit this skill has had.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-13 | a792ab0e17bb | CAUTION | B | 89 | first audit |
Questions
What does the Debug skill do?
The Claude Code background agent for Linear, Slack, Github, GitLab etc. you deploy anywhere. Supports Codex, Cursor, Gemini, and Opencode harnesses too.
Is Debug safe to install?
With care. The audit graded it B (89/100) and found 5 things worth knowing before you trust this skill, listed below with the exact line each was found on.
What can Debug access on my machine?
The audit observed no filesystem, network or shell use at all in its source.
How current is this page?
The grade is for one exact copy of the source (a792ab0e17bb), read on 2026-10-13. The repository is watched, and a new audit runs when it changes — this is the first audit.