feat(mcp): migrate to mcp 2.x (MCPServer), replacing the <2 ceiling

Pilot for moving the fleet off mcp 1.x before it goes away. The port is
three mechanical changes plus docs:

  - import:      mcp.server.fastmcp.FastMCP -> mcp.server.mcpserver.MCPServer
  - constructor: transport options moved off it; MCPServer(name) only
  - run():       host/port/stateless_http/transport_security are now kwargs,
                 and `mcp.settings` no longer exists

`@mcp.tool()` is unchanged — all six tools keep their decorators and
signatures untouched.

Verified, not assumed:

1. tools/list is byte-for-byte identical between mcp 1.27.1 and 2.x
   (6 tools, same names, same inputSchema properties and required lists,
   dumped with by_alias=True so the comparison is in wire format). This
   was the real risk: 2.x switched the protocol types to snake_case
   internally, and tool schemas are what decide routing.
2. The server boots for real over streamable-http with stateless_http=True
   and DNS-rebinding protection disabled, and answers initialize 200 with
   the same capability set.
3. CI now runs `import docs_mcp.server` inside the built image before push.

One behavior change: an unversioned server reports serverInfo.version ""
under 2.x instead of the SDK version, so Drawbar will see an empty version
string. Cosmetic — left as-is rather than inventing a version scheme.

Dependency note: 2.x has no [fastmcp] extra, replaces httpx with httpx2,
and requires opentelemetry-api. Our own httpx>=0.27 pin is unaffected —
httpx2 is a separate distribution.

Co-Authored-By: Claude Opus 5 (1M context) <[email protected]>
Claude-Session: https://claude.ai/code/session_01FnVuG79cYPcRLTp4pC8ujR

CLAUDE.md

@@ -161,7 +161,7 @@ template's PLAN.md.
 │       ├── becks_pfr.py
 │       └── becks_products.py
 ├── rag/                          # chunk + embed + Chroma + BM25
-├── docs_mcp/                     # FastMCP server + lessons.md
+├── docs_mcp/                     # MCPServer (mcp 2.x) + lessons.md
 ├── eval/                         # Golden-query harness
 └── scripts/                      # registry_gc.py, usage_report.py
 ```
@@ -189,7 +189,8 @@ template's PLAN.md.
 - **Lexical store**: SQLite FTS5
 - **Fusion**: RRF k=60
 - **Transport**: streamable-HTTP in prod, stdio for local dev
-- **MCP framework**: FastMCP with `stateless_http=True`
+- **MCP framework**: `mcp` 2.x `MCPServer` (the 1.x `FastMCP`), with
+  `stateless_http=True` passed to `run()`
 
 ### Image name and package linking are repo-name-derived
 
@@ -230,7 +231,9 @@ python -m eval.run_eval --queries eval/queries.jsonl --output eval/results/basel
 - **Reranker per-pair token limit**: jina-reranker GGUF rejects the
   ENTIRE batch if any doc exceeds `n_ctx_train=1024`. Truncate
   reranked docs to ~2000 chars.
-- **FastMCP `stateless_http=True`**: critical for prod.
+- **`stateless_http=True`**: critical for prod. Under mcp 2.x it is a
+  `run()` kwarg, NOT a constructor arg — every transport option moved off
+  the constructor. `mcp.settings` no longer exists.
 - **Runner shell is `/bin/sh` (dash)** in CI — no `${VAR::N}`.
 - **Cloudflare 100 MB body cap**: push via LAN endpoint
   `192.168.0.2:1234`, pull via `git.jpaul.io`.

README.md

@@ -147,7 +147,7 @@ PRODUCT_NAME=crop_seed python -m docs_mcp.server \
 │   ├── index.py                   # Chroma + BM25 builder
 │   └── bm25.py                    # FTS5 lexical index w/ seed-domain facets
 ├── docs_mcp/
-│   ├── server.py                  # FastMCP — 6 tools, hybrid+rerank
+│   ├── server.py                  # MCPServer — 6 tools, hybrid+rerank
 │   ├── lessons.md                 # Curated knowledge layer (Pioneer fallback)
 │   └── usage.py                   # TimedCall + JSONL telemetry
 ├── eval/

docs_mcp/server.py

@@ -25,7 +25,8 @@ import re
 from pathlib import Path
 from typing import Annotated, Any
 
-from mcp.server.fastmcp import FastMCP
+from mcp.server.mcpserver import MCPServer
+from mcp.server.transport_security import TransportSecuritySettings
 from pydantic import Field
 
 from .usage import TimedCall
@@ -57,9 +58,12 @@ RRF_K = int(os.environ.get("RRF_K", "60"))
 
 
 # ---------------------------------------------------------------------------
-# FastMCP setup.
+# MCPServer setup.
+#
+# mcp 2.x moved every transport option (stateless_http, host, port,
+# transport_security, ...) off the constructor and onto run() — see main().
 # ---------------------------------------------------------------------------
-mcp = FastMCP(f"{PRODUCT_NAME}-docs", stateless_http=True)
+mcp = MCPServer(f"{PRODUCT_NAME}-docs")
 
 
 # ---------------------------------------------------------------------------
@@ -1196,12 +1200,20 @@ def main() -> None:
 
     if args.transport == "stdio":
         mcp.run()
-    else:
-        mcp.settings.host = args.host
-        mcp.settings.port = args.port
+        return
+
+    # mcp 2.x: transport options are run() kwargs, and `mcp.settings` is gone.
+    run_kwargs: dict[str, Any] = {"host": args.host, "port": args.port}
+    if args.transport == "streamable-http":
+        # Was a constructor arg under 1.x. Required in prod — see CLAUDE.md.
+        run_kwargs["stateless_http"] = True
     if os.environ.get("MCP_DISABLE_DNS_REBINDING_PROTECTION") in {"1", "true", "yes"}:
-            mcp.settings.transport_security.enable_dns_rebinding_protection = False
-        mcp.run(transport=args.transport)
+        # We're only reachable as `seed-mcp:8080` on Drawbar's internal docker
+        # network, never published to a host port.
+        run_kwargs["transport_security"] = TransportSecuritySettings(
+            enable_dns_rebinding_protection=False,
+        )
+    mcp.run(transport=args.transport, **run_kwargs)
 
 
 if __name__ == "__main__":

requirements.txt

@@ -1,5 +1,5 @@
 # MCP server
-mcp[fastmcp]>=1.0.0,<2  # 2.x removed mcp.server.fastmcp (FastMCP -> MCPServer); see docs_mcp/server.py
+mcp>=2,<3          # 2.x: no [fastmcp] extra; FastMCP -> MCPServer
 pydantic>=2.0
 httpx>=0.27