Signet Eval MCP Server

Deterministic policy enforcement for AI agent tool calls. Integrated tool calls pass through user-defined rules before execution. No LLM in the authorization path. Advisory nudges are separate from authorization.

People who need security tools inside Claude, Cursor, VS Code, or another MCP client. The project is written in Rust.

VERIFIED ACTIVE

LAST COMMIT 2026-09-18 · ★ 3 · #180 OF 204 MAINTAINED SECURITY · VERIFIED 2026-09-18

MIT · Rust servers · how we verify → /methodology

01 · Evidence

Security posture

What to check before giving this server access to your agent - from the registry, GitHub, and our own probes. We don't score safety; we show what's verifiable.

license MIT - declared in the repository

registry namespace io.github.jmcentire is GitHub-verified and matches the repo owner

02 · What Signet Eval can do

Prose above is summarized from the project's README and registry record - no invented capabilities.

Latest releases

v3.12.2 · 2026-08-28

Antigravity permissions-adapter repair. · Consumes Antigravity's native nested toolCall payload and emits native top-level allow/deny decisions. · Normalizes command, file, multi-edit, and MCP calls into Signet's…

v3.12.1 · 2026-08-01

Durable safety-default release. · Ships the GitHub remote-owner identity guard and embeds its check for atomic first-use installation. · Restores hard denial of ephemeral Task state in favor of Kindex tasks. · Requires…

v3.12.0 · 2026-06-13

Add Antigravity and OpenCode hook adapters, and change prefer_persistent_task_store default policy to Ask

03 · Who maintains Signet Eval

signet-eval is maintained by jmcentire. It's the only MCP server we track from this author; the repo dates to Mar 2026.

04 · Facts

category
security - ranked #180 of 204 actively-maintained security servers as of 2026-09-18.
release cadence
2 releases in the last 90 days (latest 2026-08-28)
registry
io.github.jmcentire/signet-eval (active, first published 2026-05-16 · 2 versions)

05 · Signet Eval FAQ

What is Signet Eval?

Deterministic policy enforcement for AI agent tool calls. Integrated tool calls pass through user-defined rules before execution. No LLM in the authorization path. Advisory nudges are separate from authorization.

Is Signet Eval still maintained?

Yes - as of 2026-09-18, its last commit was 2026-09-18 and it shipped 2 releases in the last 90 days. We re-verify nightly.

06 · Alternatives to Signet Eval