MCP server — intent security pre-flight checks for autonomous AI agents
Drift inferred · capture-to-capture
No drift recorded — single capability capture; advisories appear once its surface changes.
transport stdio
verified
reported
listed in the official MCP registry counts 4 tools · 0 res
· 0 prompts
permission surface via code analysis
tools
-
check_intent
-
expand_intent
-
list_actions
-
review_action
evidence-backed
findings quoted directly from the published source artifact — not inferred
code files: 12
network 4
- net package/dist/tools/check-intent.js :24
const response = await fetch(`${API_URL}/v1/check`, { - net package/dist/tools/expand-intent.js :17
const response = await fetch(`${API_URL}/v1/expand`, { - net package/dist/tools/list-actions.js :7
const response = await fetch(`${API_URL}/v1/actions`, { - net package/dist/tools/review-action.js :18
const response = await fetch(`${API_URL}/v1/review`, {
tool registrations 4
- check_intent package/dist/tools/check-intent.js :4
- expand_intent package/dist/tools/expand-intent.js :4
- list_actions package/dist/tools/list-actions.js :3
- review_action package/dist/tools/review-action.js :4
install hooks 1
- prepublishOnly package/package.json :24
npm run build
declared dependencies 5
- @modelcontextprotocol/sdk@^1.27.0
- zod@^3.25.0
- @types/node@^20.0.0
- typescript@^5.7.0
- vitest@^3.0.0