Exposes selected operations of an existing REST/OpenAPI app as MCP tools with generated schemas, permission annotations, and a risk class, requiring explicit opt-in before anything is listed and a second flag for destructive calls. Enforces security-first guardrails such as environment-injected auth, a single confined base URL, timeouts, response caps, rate limiting, and a local test console for trying tools.
Enables read-only MCP access to Yandex Tracker for retrieving issues with comments and attachment metadata, searching issues, reading issue history, downloading attachments, and finding users. It also supports connection status checks and token configuration without exposing secrets.
Enables agents to join shared teams and run project work—expanding projects into features and tasks, claiming items, tracking priorities and dependencies, linking delivery versions to Git commits or artifact hashes, and recording validation gates with evidence—through 23 tools that use the same REST commands and permissions as the human dashboard. Tokens bind each agent to its own identity, so it cannot impersonate a PM, overwrite another member's claim, or approve its own delivery.
Enables LLM agents to discover and invoke guarded web search, read-only SQLite queries, and sandboxed file read/write tools over the Model Context Protocol. It adds schema validation, rate limiting, output sanitization, request-ID logging, and Prometheus metrics at the protocol boundary.