AgentKitBLOCK
Governed MCP server and agent runtime: signed action approvals, typed tool effects and audit logs for Claude Desktop, Cursor and custom agents.
Overview
From the repository's own README, as read at the audited commit. Badges and raw HTML are left out.
[](https://github.com/Yacine-ai-tech/AgentKit/actions/workflows/ci.yml) [](LICENSE) [](https://glama.ai/mcp/servers/Yacine-ai-tech/AgentKit) [](https://agentkit.ysiddo-ai-projects.app) [](https://agentkit.ysiddo-ai-projects.app/research) [](https://agentkit.ysiddo-ai-projects.app/benchmark) [](https://agentkit.ysiddo-ai-projects.app/guide)
An MCP server where tools are declarative, effects are typed, and every action is policy-gated and audited — usable by any MCP client (Claude Desktop, Cursor, LangGraph, Claude Agent SDK, CrewAI).
Live Application: agentkit.ysiddo-ai-projects.app.
- Research Background:
RESEARCH.md/ Online Research Documentation - Empirical Benchmarks:
BENCHMARK.md/ Online Benchmark Dashboard - User Guide: Online User Guide
- Self-Hosting Guide:
SELF_HOSTING.md
Three things distinguish it from a typical MCP server:
- Declarative tools — define tools in YAML over your own Postgres or HTTP API. No
Python, no fork. (docs/REUSE.md)
- Real actions, not just reads — tools declare an effect (
read/write/
`destructi
adeb37293633OBSERVED · 2026-10-13Connect
Built from this server's own package name, version and transport as found in its source — not copied from anyone's documentation, so it cannot drift against a page we do not control. Replace the environment placeholders with a token scoped to the least it needs.
claude mcp add agentkit-frontend --env ADMIN_PASS=${ADMIN_PASS} --env ADMIN_TOKEN=${ADMIN_TOKEN} --env AGENTKIT_APPROVAL_TOKEN=${AGENTKIT_APPROVAL_TOKEN} --env AGENTKIT_INTERNAL_TOKEN=${AGENTKIT_INTERNAL_TOKEN} -- npx -y [email protected]{
"mcpServers": {
"agentkit-frontend": {
"command": "npx",
"args": [
"-y",
"[email protected]"
],
"env": {
"ADMIN_PASS": "${ADMIN_PASS}",
"ADMIN_TOKEN": "${ADMIN_TOKEN}",
"AGENTKIT_APPROVAL_TOKEN": "${AGENTKIT_APPROVAL_TOKEN}",
"AGENTKIT_INTERNAL_TOKEN": "${AGENTKIT_INTERNAL_TOKEN}"
}
}
}
}Exposed tools (1)
1 read · 0 write · 0 destructive.
| Tool | Risk | Description |
|---|---|---|
t | read | d |
Trust audit
BLOCKgrade D · trust 69/100 Do not install this without reading the findings. The audit found something that could harm you or your machine.
| Layer | What it checks | Result |
|---|---|---|
| L0 | Provenance & inventory | PASS |
| L1 | Static analysis of the code | FAIL |
| L2 | Instruction surface (what it tells the agent) | PASS |
| L3 | Class-specific surface | PASS |
| L4 | Behavioural (sandbox) | SKIPPED |
What the source does
- Filesystem
- none-observed
- Network
- declared (2 observation(s))
- Shell
- declared (1 observation(s))
- Dependencies
- not all pinned
- Secrets in source
- none-found
Findings (11)
import{a as Li,i as Qw,j as K,P as eO,S as Ib,D as tO,b as kb,C as Er,E as Ba,c as xn,L as rO}from"./index-DIROj1LC.js";import{V as Si,g as ce,a as D,U as S,s as nO,Y as iO,_ as aO,u as Kh,$ as oO}fro*/function Cr(){return Cr=Object.assign?Object.assign.bind():function(o){for(var c=1;c<arguments.length;c++){var a=arguments[c];for(var p in a)({}).hasOwnProperty.call(a,p)&&(o[p]=a[p])}return o},Cr.a# Your own GPU box on the LAN ... http://192.168.1.50:11434
.flake8
.env.production
mod = importlib.import_module(module_name)
ToolPolicy(name="bad", effect="sudo")
@fontsource/inter, framer-motion, lucide-react, react, react-dom, react-router-dom, recharts, @playwright/test
fastmcp, fastapi, langgraph, langchain, litellm, anthropic, groq, openai
frontend/dist/logo_agentkit.png
frontend/public/logo_agentkit.png
Gates applied: no_behavioural_pass.
adeb37293633full audit observations/trust-audit/mcp-server/yacine-ai-tech__agentkit.json · Report an issue / request a re-scanAudit history
Every audit this server has had. A grade with a past is a grade somebody is still checking.
| Date | Source | Verdict | Grade | Score | Change |
|---|---|---|---|---|---|
| 2026-10-13 | adeb37293633 | BLOCK | D | 69 | first audit |
Questions
What is the AgentKit MCP server?
Governed MCP server and agent runtime: signed action approvals, typed tool effects and audit logs for Claude Desktop, Cursor and custom agents.
What tools does AgentKit expose?
1 in total: 1 read-only, 0 that write, and 0 that can delete or overwrite. Every one is listed on this page with its risk.
Is AgentKit safe to connect to an agent?
No — not without reading the findings first. The audit graded it D (69/100) and found 1 critical or high issue in the source. Each one is listed on this page with the file and line it is on.
What credentials does AgentKit need?
It reads ADMIN_PASS, ADMIN_TOKEN, AGENTKIT_APPROVAL_TOKEN, AGENTKIT_INTERNAL_TOKEN, ANTHROPIC_API_KEY, DSPY_MAX_TOKENS, GROQ_API_KEY, INFERENCE_TOKEN, INTERNAL_TOKEN, LLM_TOKEN, LOCAL_LLM_TOKEN and MCP_AUTH_TOKEN from the environment. Give it a token scoped to the least it needs — an agent that can be talked into calling a tool can be talked into calling it with your credentials.
How does AgentKit run?
It speaks sse and stdio, so it runs as a local process your client starts. It is published on npm as agentkit-frontend at 1.0.0.
How current is this page?
The grade is for one exact copy of the source (adeb37293633), read on 2026-10-13. The repository is watched and re-audited when it changes.