Skip to content

signetry-claude-code

License PRs Welcome

Govern Claude Code's changes in real time — with signed receipts.

The Claude Code plugin for signetry-core: Signetry decides how much authority an agent's change has earned and proves it. This plugin brings that governance inside Claude Code — enforced by deterministic code, never by the model itself (an agent can't approve its own change).

Part of the Signetry platform.

Prerequisite: pip install "signetry-core @ git+https://github.com/Signetry/core@v0.8.0" and a .signetry/admission.yaml in your repo (a conservative default applies without one).

What it does

A PreToolUse hook runs before every Edit / Write / Bash and blocks out-of-scope or forbidden actions before they happen — using signetry guard (deterministic, not the model). It bundles the Signetry MCP server and an /signetry:admit skill for on-demand full admission with a signed receipt.

  • Agent tries to edit deploy.yml / .env / a secret, or run curl … | bash / git pushblocked with a reason, before it happens.
  • Agent edits an in-scope file → allowed silently.
  • /signetry:admit → runs the full admission pipeline and returns an Admission Decision Pack + signed receipt.

Install

/plugin marketplace add Signetry/claude-code
/plugin install signetry@signetry-claude-code

Or test locally: claude --plugin-dir ./signetry

Layout

  • signetry/.claude-plugin/plugin.json — plugin manifest
  • signetry/hooks/ — the PreToolUse guard + session-start hooks
  • signetry/scripts/signetry-mcp.sh — launches the Signetry MCP server
  • signetry/skills/admit/SKILL.md — the /signetry:admit skill
  • signetry/.mcp.json — MCP server registration

Scan & fix (from the same signetry-core)

Beyond the editor guard, signetry-core can find vulnerabilities and govern the fix:

signetry scan .                 # SAST over the repo (7 languages, SARIF), offline & free
signetry scan . --fix --fix-agent claude-code   # draft a governed fix → admission → signed receipt

--fix is bring-your-own-key (your ANTHROPIC_API_KEY, never shared, redacted from every artifact) and opens branch-only PRs — never merges. See signetry-core: AUTOFIX_SETUP.md.

Guarantees

  • Fail-closed on admit; the soft guard may fail open only with a loud INACTIVE.
  • Never auto-merges. auto_merge is always false — a human merges.
  • Governance logic lives in signetry-core; this plugin never reimplements policy.

See SECURITY.md · PRIVACY.md · umbrella overview.

License

Apache-2.0. Use it, fork it, ship it commercially — no strings.

This repository is part of Signetry's open-core model: the integration surface is Apache-2.0 so anyone can add an agent, an editor, or a CI adapter, while the engine (Signetry/core) is source-available under BUSL-1.1 and converts to Apache-2.0 on 2030-08-31.

Contributions are accepted under the CLA — it lets us move a well-built adapter into the engine later without asking every contributor for permission again.

About

Claude Code plugin for Signetry: a PreToolUse guard that blocks out-of-scope/forbidden agent actions before they happen, plus the Signetry MCP server and /signetry:admit skill. Split from signetry-plugins.

Topics

Resources

Code of conduct

Contributing

Security policy

Stars

0 stars

Watchers

0 watching

Forks

Releases

Packages

Used by

Contributors

Languages