MCP Risk — AI view

This page is the machine-first rendering of the MCP Risk website. Same content, no presentation. Alternate surfaces: /llms.txt (plain-text orientation), /api/site (structured JSON).

Summary

Product
Evidence-backed trust profiles for MCP repos and configs. Approve a known-good state and monitor what changes.
Headline
Approval has a shelf life.
Positioning
Servers get patched, descriptions get rewritten, and tools appear that nobody signed off on. We hold the surface you approved and measure everything after it against that.
Capabilities
Approval records; Config fingerprints; Drift evidence; Trust card gallery

New to MCP?

The Model Context Protocol (MCP) lets AI agents connect to external tools and data through MCP servers — local processes or remote services that expose actions like reading files, querying databases, or calling APIs.

That power is the risk. An MCP server can ship a tool whose description quietly instructs the agent to exfiltrate secrets, or change behavior after you approve it. MCP Risk reads the repo and config, produces a trust profile you can approve, and flags when the approved state drifts.

Method

A trust profile is not a scan. It is an approval record. The scanner only produces evidence. The product is the lifecycle around that evidence: who approved what, under which config, and what changed afterwards.

  1. Evidence: Extract declared tools, install commands, dependencies, env keys, permission surfaces, and risky code patterns.
  2. Config: Evaluate the actual launch context: command, args, tokens, filesystem scope, network exposure, and client.
  3. Approval: Bind decisions to repo version, config hash, policy version, analyzer version, reviewer, and environment.
  4. Drift: Trigger re-review when tools, descriptions, dependencies, env vars, or install commands change materially.

Review flow example

Every MCP approval should leave evidence behind. A calm chain from tool request, to trust profile, to approval, to drift.

$ mcp-risk profile github.com/example/mcp-server
subject: github.com/example/mcp-server
config: 9f1c…42ab
verdict: approve_with_conditions

$ mcp-risk diff approved latest
change: env_key_added GITHUB_TOKEN
change: tool_added run_shell
result: re-review required

Approved profile vs. the live server, three weeks later.

  subject: github.com/example/mcp-server
  config: 9f1c…42ab
+ env_key_added: GITHUB_TOKEN
+ tool_added: run_shell
- verdict: approve
+ verdict: re-review required

Approval lapsed — re-review required before this server runs again.

API for agents

Submit a repo or config. See the verdict and summary instantly — add your email to unlock the full evidence and downloadable card.

POST /api/submit
Body: {"repoOrConfig": "<github url or mcp.json contents>"}. Returns a profile URL plus queued or completed scan status, verdict, risk level, summary, and public slug. No authentication required.
GET /api/profile/{publicSlug}/status
Returns the current scan status and profile while asynchronous scanner evidence is produced.
GET /api/site
This site's full content as structured JSON.
GET /llms.txt
Plain-text orientation document following the llms.txt convention.