peopleanalyst

Platform · Shipped · new here

MCP agent governance — keys, scopes, rate-limiting, sensitive-op confirmation, audit

AI-native control plane for the MCP gateway: per-consumer keys to scope sets, enforced per-consumer rate-limiting (token bucket), sensitive-op confirmation (a mutating tool requires confirm:true or is rejected + audited), and a full audit trail. The Barndoor/Aegis governance primitive set, in-house.

Cluster
Platform
Type
service
Status
Shipped
Used by
people-analytics-toolbox, performix

Maturity evidence

Functioning
Live in prod 2026-07-01: per-consumer token-bucket rate-limit (AGENTGOV-A) + sensitive-op confirmation gate (AGENTGOV-B; first adopter job-family-agent.match.review) atop per-consumer keys/scopes + the mcp_audit trail.
Valuable
Governs AI agents operating our tools without buying Barndoor/Aegis: throttle abuse, require confirmation for destructive ops, attribute + audit every call.
Understood
docs/DECISIONS/2026-06-29-... (buy/build) + AGENTS.md MCP transport section; src/lib/mcp/ (auth/scope/rate-limit/wrap/audit).
Integrated
Enforced in wrapTool for every MCP tool; env kill-switch/tuning for rate limits; sensitive tools opt in via sensitive:true.

How to use it

MCP
*
Data
src/lib/mcp/

Other Platform capabilities

← Every platform capability

Capability detail sourced from the People Analytics Toolbox capability feed (source of truth), snapshot retrieved 2026-07-23.