ai-architect-mcp-spec

# AI Architect MCP Spec (v0.7.0) — Dispatcher Protocol You (the host) drive a loop: 1. Call `start_pipeline` → receive an envelope with `{ run_id, messages, action, ... }`. 2. Display every entry in `messages` to the user. 3. Execute `action` per the dispatch table below. 4. Call `submit_action_result(run_id, result)` → receive next envelope. 5. Repeat until `action.kind === "done"` or `action.kind === "failed"`. **`emit_message` is never returned to the host as `action`.** The runner coalesces all status messages into the `messages` array; the `action` field always carries something the host actually has to do. --- ## ENVELOPE SHAPE Every response from `start_pipeline` and `submit_action_result` has this shape: ```json { "run_id": "run_abc_123", "current_step": "context_detection", "messages": [ { "text": "🟢 PRD Spec Generator — TRIAL TIER\n...", "level": "info" } ], "action": { "kind": "ask_user" | "call_pipeline_tool" | ... , ... }, "state_summary": { "sections": [...], "clarification_rounds": 3, "errors": 0 } } ``` - `run_id` — handle for `submit_action_result` and `get_pipeline_state`. Caputre once from the first response and reuse. - `messages` — banners/status lines collected while the runner advanced internally to reach `action`. May be empty. Display each `text` at the given `level` (default `info`) before executing `action`. - `action` — what you must execute. **Never `emit_message`.** Always one of: `ask_user`, `call_pipeline_tool`, `call_c...