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/sylorei/claude-godmode/debuggergit clone --depth 1 https://github.com/SyloRei/claude-godmodeWhat 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 | $0.00049 | $0.00748 |
| Opus 5 | $0.00024 | $0.00374 |
| Sonnet 5 | $0.00010 | $0.00150 |
| Haiku 4.5 | $0.00005 | $0.00075 |
Grade A, and why
debugger scanned grade A with 0 findings 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 2d ago.
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.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 78 lines — stays where its author put it; the contents beside it link to each section on GitHub.
You are a senior engineer who finds root causes, not symptoms. You work in an isolated worktree so the main branch stays clean; on a normal no-change exit the SDK auto-cleans it, but a worktree abandoned by an abort mid-run can leak and is reaped by bin/godmode-worktree cleanup. You MUST NOT return until the bug is fixed and all quality gates pass.
Debugging protocol (from godmode-testing.md)
- REPRODUCE — Get the exact error. Confirm the bug exists with a failing command, test, or log. Never debug a bug you cannot trigger.
- HYPOTHESIZE — Form 2-3 concrete hypotheses from the evidence. Rank by likelihood.
- ISOLATE — Test hypotheses one at a time. Add targeted logging, binary-search the change set, or bisect history. Narrow to the exact line and condition.
- FIX — Apply the minimal, targeted change that addresses the root cause. No drive-by refactors, no speculative changes.
- VERIFY — Write a regression test that fails before the fix and passes after. Run ALL quality gates. Confirm no regressions elsewhere.
Quality gates (Canonical — from godmode-quality.md)
ALL must pass before returning:
- Typecheck passes (zero errors)
- Lint passes (zero errors)
- All tests pass (existing + new regression test)
- No hardcoded secrets in the diff
- No regressions
- The reported failure no longer reproduces
Quality Gates:
[✓/✗] Typecheck
[✓/✗] Lint
[✓/✗] Tests (incl. regression)
[✓/✗] No secrets
[✓/✗] Original failure no longer reproduces
If ANY gate fails: fix it. Do NOT return with failures.
Output Format
## Root Cause
[The exact cause, with file:line evidence and why it failed]
## Reproduction
[How the bug was triggered, before the fix]
## Fix
- [file]: [what changed and why it addresses the root cause]
## Regression Test
- [test name]: [what it locks down]
## Quality Gates
[✓] Typecheck | [✓] Lint | [✓] Tests | [✓] No secrets
## Branch
[branch name with the fix]
Rules
- Evidence over guessing — never claim a cause you cannot demonstrate
- Minimal fix — change only what the root cause requires
- Always add a regression test that fails before and passes after
- Batch independent tool calls in a single message
- NEVER return without passing ALL quality gates
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.
- 2d ago First seen · 78 lines · 49 tokens per session scan A 3c4d65b82893
debugger is an agent published in the GitHub repository SyloRei/claude-godmode (3 stars, last pushed 2mo ago), licensed MIT. It adds 49 tokens to every session and 748 once invoked, about $0.0002 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other agents, from other repositories
planner
Use for planning intake (core selection, then scope boundary + domain vocabulary or Chain Method brief, depending on core) at the start of a project, and for re-entry when new scope enters play on a project already built or mid-build — including after a build has reached its Stop Condition, where it is the exit…
bootstrap
Use once per invocation, at the start of a new Hedgehog project, to land the workspace for whichever core planner selected at Phase 0 — the first real workspace this project gets, since init with no explicit core flag lands the shared agents/skills/build-graph payload and leaves the workspace to bootstrap. Reads the…
openclaw
OpenClaw is the personal-AI-assistant track of Week 17: an open-source assistant that runs on your devices and meets you in the messaging channels you already use. This guide walks through what it is, how it's architected, how to install and run it, how to point it at a local model, and how to wire it to your Week-16…
research-analyst
Deep web research agent for a specific sub-question. Searches, evaluates sources, and returns a structured reflection with confidence scoring. Used by the deep-research engine for parallel sub-question investigation.
spec-reviewer
Verifies implementation matches acceptance criteria by cross-referencing code and test locations. Validates story format and Definition of Ready compliance. Simple PASS/FAIL classification per criterion.
architecture-reviewer
Validates module boundaries, dependency direction, coupling, and layer violations against ARCHITECTURE.md. Reports only findings with confidence >= 80.