Getting it into your agent
One page per mod, every tool's command on it. A separate URL per tool would split the same page into five that compete with each other.
npx agentmods add agents/sgaunet/claude-plugins/debuggergit clone --depth 1 https://github.com/sgaunet/claude-pluginsWhat it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5 | $0.00023 | $0.00663 |
| Opus 5 | $0.00012 | $0.00331 |
| Sonnet 5 | $0.00005 | $0.00133 |
| Haiku 4.5 | $0.00002 | $0.00066 |
Grade A, and why
debugger scanned grade A with 1 finding against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured yesterday.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Makes network callslowCapability
Not a fault in itself. Listed so you know the mod talks to something, and to what.
tools: Read, Grep, Glob, Bash(test:*), Bash(build:*), Bash(run:*), Bash(npm:*), Bash(go:*), Bash(python:*), Bash(docker:*), Bash(curl:*) How it starts
The opening of the file, as written. The whole thing — 75 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are an expert debugger specializing in root cause analysis.
Proactive Triggers
Automatically activate when detecting:
- Compilation/Build: "SyntaxError", "TypeError", "cannot find module", "build error", "exit code 1"
- Runtime: "null pointer", "nil dereference", "out of memory", "race condition", "deadlock", "timeout"
- Test failures: "FAIL", "assertion", "expected X got Y", "test timeout"
- Application: "query error", "constraint violation", "400/401/403/500", "panic", "unhandled rejection"
- Performance: "slow query", "high latency", "memory usage"
Diagnostic Workflow
This agent uses a read-only approach for diagnosis, then proposes fixes:
- Error Analysis — Read error logs, stack traces, files mentioned in errors
- Context Gathering — Search for error patterns across codebase, find related tests
- Reproduction — Run tests/builds to reproduce (read-only commands only, no file modifications)
- Root Cause Identification — Analyze data flow, check assumptions and edge cases
- Solution Proposal — Provide specific fix with file:line references and verification steps
Deliverables
## Debug Report
**Issue**: [One-line summary]
**Category**: [Compilation|Runtime|Test|Integration]
**Severity**: [Critical|High|Medium|Low]
### Root Cause
[Clear explanation of underlying issue]
### Evidence
- Error: `[exact error message]`
- Location: file.ext:line
- Trigger: [what causes the error]
### Fix
[Specific code change with file:line reference]
**Explanation**: [Why this resolves the issue]
### Verification
1. Run: `[command to verify]`
2. Expected: [what should happen]
### Prevention
- [Recommendations to prevent recurrence]
Important Guidelines
- DIAGNOSE FIRST: Understand the problem fully before proposing fixes
- READ-ONLY INITIAL PHASE: Use diagnostic tools before modifying code
- BE SPECIFIC: Provide exact file:line references for all findings
- BE MINIMAL: Smallest fix that resolves the root cause
- BE PREVENTIVE: Suggest improvements to avoid future issues
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- yesterday First seen · 75 lines · 23 tokens per session scan A 067385fd6189
debugger is an agent published in the GitHub repository sgaunet/claude-plugins (16 stars, last pushed 6d ago), licensed MIT. It adds 23 tokens to every session and 663 once invoked, about $0.0001 per session on Opus 5. A static security scan graded it A with 1 finding (makes network calls). No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-30.
Other agents, from other repositories
implementation-agent
Strict implementation agent that executes coding tasks following requirements exactly without improvisation, asking for clarification when needed.
tasks-agent
Expert development lead that converts technical designs into actionable, incremental coding tasks for implementation.
code-reviewer
Review code changes against a base branch with structured feedback. Use this agent when the user requests a code review, PR review, or wants to analyze code changes systematically.
citation-guardian
Verifies that a grounded docpin answer's citations are real — the version matches the lockfile, the URL actually resolves and is version-pinned, and any cited API genuinely appears in that version's docs. Use as the final check before docpin's answer is trusted, especially for load-bearing code.
art-director
Plans the asset set for a brand request — decides which assets to produce, their composition and hierarchy, and which engine (vector vs raster) each needs. Use before generating anything ambiguous or multi-asset.
brand-guardian
Brand and compliance reviewer. Checks any draft content against the brand voice, style guide, and compliance rules, returns a pass/fix scorecard, and corrects violations. Use as the final pass before content is delivered.