God Mode OS For Claude Code — Development skill for Claude Code
Discipline hooks for Claude Code.
How to install God Mode OS For Claude Code
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open Wishmakingfairy/god-mode-os-for-claude-code and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What God Mode OS For Claude Code does
Discipline hooks for Claude Code. Stops the agent from claiming 'done' without proof, blocks accidental writes to ~/.claude config, and routes prompts to skills via local pgvector with zero Anthropic tokens.
Alternatives in Development
- Done Check — PUA Done Check — 用于没跑测试别说完成、已完成但没证据、done without proof、需要验收/回归/交付质量检查的场景 19.5k ★
- Discipline — Toggle discipline mode — auto-invoke verification, brainstorming-before-coding, and review checks 2.8k ★
- Academic Pptx Skill — A Claude Skill for creating academic presentations (conference talks, seminar slides, thesis defenses, grant b 811 ★
README
Claude Code, with discipline hooks
by [@Wishmakingfairy](https://github.com/Wishmakingfairy)
**Stop Claude Code lying about "done". Stop it burning your quota on routing.**
[](LICENSE) [](CHANGELOG.md) [](#compatibility) [](https://github.com/Wishmakingfairy/god-mode-os-for-claude-code/actions/workflows/ci.yml)

god-mode-os is a layer of bash hooks for Claude Code that:
- Block false "done" claims. Claude can't say it ran the tests if it never ran the tests.
- Protect
~/.claude/. Writes to your hooks, skills, andsettings.jsonneed explicit approval. - Catch committed secrets before you ship. A local, fail-closed gitleaks gate.
- Route prompts to skills locally. pgvector + Ollama, sub-300 ms, zero Anthropic tokens.
- Run offline. No telemetry, no phone-home. Your code and prompts never leave the machine.
git clone https://github.com/Wishmakingfairy/god-mode-os-for-claude-code
cd god-mode-os-for-claude-code
./install.sh # ~60 seconds, only jq + python3 needed
**Don't trust the README?** Run `bash smoke.sh`. It tests every load-bearing claim in this README against the real code and prints a pass/fail receipt. It runs in a throwaway `HOME` and installs nothing, so you can prove the claims before you touch your own `~/.claude`. The same script runs in CI on every push, on Linux and
Related Skills
Proof Of Done
Claude Code plugin that stops a coding agent from ending its turn with 'tests pass', 'build succeeds' or 'depl
Mythos Router
The leaked Anthropic reasoning protocol. Running locally. Zero-drift coding with Strict Write Discipline and a
Harness Verify
Run the full verification loop before claiming any feature is complete. This is the "Ralph Wiggum Loop" - forc
Hunt Takeovers
Enumerate subdomain-takeover candidates for an ingested program (passive subfinder + dnsx CNAME + subzy finger
Rolepod Stats
Show rolepod evidence stats for this project — tier routes, verify/review verdicts, strong-dispatch overrides,
Did You Check
Checks whether the Claude Code best practices were actually followed, and blocks the turn when they were not:
Related Agents
Verify Complexity
STAGE 1 VERIFICATION - Fast complexity check. Flags monster files (>1000 LOC), high cyclomatic complexity (>15
Testing Reality Checker
Stops fantasy approvals, evidence-based certification - Default to "NEEDS WORK", requires overwhelming proof f
Grok Agent 01
ONLY job: find free or freemium AI token/API plans (OpenAI-compatible, Anthropic, Google, open routers, studen