open-coder-ai/chock
Governance-as-code for AI coding agents: author a policy once, enforce it on every agent (Claude, Copilot, Cursor, Codex) via git hooks, CI, and native controls.
ARCHETYPE
Skill Collector
Ten-plus skills loaded on demand. A procedural-knowledge library.
Copy this rig
# review before running: this installs third-party code
$ npx degit open-coder-ai/chock/.claude ./rig-chock # inspect, then merge into .claude/ MCP servers are added to Claude Code at local scope; env vars are shown as YOUR_… placeholders — we never store values. Files are fetched with degit into a separate folder so you can review before merging.
$ npx degit open-coder-ai/chock/.agents/policies/agent-discipline/skills/agent-discipline .claude/skills/agent-discipline $ npx degit open-coder-ai/chock/.agents/policies/block-destructive-commands/skills/block-destructive-commands .claude/skills/block-destructive-commands $ npx degit open-coder-ai/chock/.agents/policies/block-invisible-unicode/skills/block-invisible-unicode .claude/skills/block-invisible-unicode $ npx degit open-coder-ai/chock/.agents/policies/block-no-verify/skills/block-no-verify .claude/skills/block-no-verify $ npx degit open-coder-ai/chock/.agents/policies/block-wildcard-agent-permissions/skills/block-wildcard-agent-permissions .claude/skills/block-wildcard-agent-permissions $ npx degit open-coder-ai/chock/.agents/policies/code-safety/skills/code-safety .claude/skills/code-safety $ npx degit open-coder-ai/chock/.agents/policies/context-hygiene/skills/context-hygiene .claude/skills/context-hygiene $ npx degit open-coder-ai/chock/.agents/policies/git-safety/skills/git-safety .claude/skills/git-safety $ npx degit open-coder-ai/chock/.agents/policies/injection-defense/skills/injection-defense .claude/skills/injection-defense $ npx degit open-coder-ai/chock/.agents/policies/memory-discipline/skills/memory-discipline .claude/skills/memory-discipline $ npx degit open-coder-ai/chock/.agents/policies/pin-github-actions/skills/pin-github-actions .claude/skills/pin-github-actions $ npx degit open-coder-ai/chock/.agents/policies/protect-agent-config/skills/protect-agent-config .claude/skills/protect-agent-config $ npx degit open-coder-ai/chock/.agents/policies/protect-commit-privacy/skills/protect-commit-privacy .claude/skills/protect-commit-privacy $ npx degit open-coder-ai/chock/.agents/policies/protect-main-branch/skills/protect-main-branch .claude/skills/protect-main-branch $ npx degit open-coder-ai/chock/.agents/policies/scan-secrets/skills/scan-secrets .claude/skills/scan-secrets $ npx degit open-coder-ai/chock/.agents/policies/token-efficiency/skills/token-efficiency .claude/skills/token-efficiency $ npx degit open-coder-ai/chock/.agents/policies/verify-dependency-exists/skills/verify-dependency-exists .claude/skills/verify-dependency-exists $ npx degit open-coder-ai/chock/.agents/skills/chock-init .claude/skills/chock-init $ npx degit open-coder-ai/chock/.agents/skills/eval .claude/skills/eval $ npx degit open-coder-ai/chock/.agents/skills/optimize .claude/skills/optimize $ npx degit open-coder-ai/chock/.agents/skills/policy-init .claude/skills/policy-init $ npx degit open-coder-ai/chock/.agents/skills/validate .claude/skills/validate
Merge into .claude/settings.json (project) or ~/.claude/settings.json (user). Hook commands reference scripts in the source repo — copy those too.
{
"hooks": {
"PreToolUse": [
{
"matcher": "Bash|PowerShell",
"hooks": [
{
"type": "command",
"command": "git -c \"alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh\" chock-hook .chock/bin/claude_code.py --multi \".chock/bin/index/claude_cod"
}
]
},
{
"matcher": "Write|Edit|MultiEdit|NotebookEdit",
"hooks": [
{
"type": "command",
"command": "git -c \"alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh\" chock-hook .chock/bin/claude_code.py --multi \".chock/bin/index/claude_cod"
}
]
},
{
"matcher": "*",
"hooks": [
{
"type": "command",
"command": "git -c \"alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh\" chock-hook .chock/bin/claude_code.py --multi \".chock/bin/index/claude_cod"
}
]
}
]
}
} Skills (22)
agent-disciplineblock-destructive-commandsblock-invisible-unicodeblock-no-verifyblock-wildcard-agent-permissionscode-safetycontext-hygienegit-safetyinjection-defensememory-disciplinepin-github-actionsprotect-agent-configprotect-commit-privacyprotect-main-branchscan-secretstoken-efficiencyverify-dependency-existschock-initevaloptimizepolicy-initvalidate
Hooks (11)
| event | matcher | runs |
|---|---|---|
| PreToolUse | Bash|PowerShell | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --multi ".chock/bin/index/claude_cod |
| PreToolUse | Write|Edit|MultiEdit|NotebookEdit | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --multi ".chock/bin/index/claude_cod |
| PreToolUse | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --multi ".chock/bin/index/claude_cod |
| SessionStart | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py |
| Stop | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --gate ".chock/compiled/block-invisi |
| Stop | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --gate ".chock/compiled/block-wildca |
| Stop | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --gate ".chock/compiled/pin-github-a |
| Stop | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --gate ".chock/compiled/protect-agen |
| Stop | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --gate ".chock/compiled/scan-secrets |
| PostToolUse | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --record ".chock/compiled/token-effi |
| PostToolUseFailure | * | git -c "alias.chock-hook=!test -f .chock/bin/launch.sh || { echo chock: no .chock/bin/launch.sh here, run chock sync --repo . >&2; exit 2; }; sh .chock/bin/launch.sh" chock-hook .chock/bin/claude_code.py --record ".chock/compiled/token-effi |
Similar rigs
open-coder-ai/chock-catalog
Policies that stop your coding agent from doing the thing you would have caught in review. The policy catalog for Chock: git-hook gates, command guards, and agent rules with replayed evals.
Skill Collector 7.4k tok · · Q33
EswarSk/token-lean
All-in-one <redacted> plugin for Claude Code (skill + hooks + scout subagent) with portable AGENTS.md rules for Codex, Cursor, and any coding agent.
Minimalist 454 tok · · Q0
jovesun-lab/arcgram
Arcgram — an interactive diagram engine for human–AI collaboration: an AI agent draws its reasoning as an auditable node-and-edge flow and self-audits it, delivered as one standalone HTML file. (Arcgram the AI diagram tool — not Archigram.)
Minimalist 31 tok · · Q0
vitaliikapliuk/modelharness
Make every model cheaper or better. Zero-config behavioral harness for Claude Code, with a reproducible 408-run benchmark
Pragmatist 236 tok · · Q0