fix(mcp): reclaim stale stdio bridge clients

CHANGELOG.md

@@ -7,6 +7,12 @@ adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
 ## [Unreleased]
 
 ### Added
+- **MCP client diagnostics and safe disconnects.**
+  `MCPSocketServer.clientConnections` reports a private connection id, the
+  kernel-reported peer pid when available, connection time, and last byte
+  activity — never a command line or path. `onConnectionsChange` publishes the
+  same snapshots, and `disconnectClient(id:)` closes exactly that socket
+  without signalling or killing its process.
 - **`SourceCursorStore.save(changed:all:)`** — the call the periodic cursor
   save makes. `changed` is what actually moved; `all` is the complete set, so
   a store that can only replace still has what it needs. The default writes
@@ -19,6 +25,16 @@ adheres to [Semantic Versioning](https://semver.org/spec/v2.0.0.html).
   makes the claim below assertable rather than timed.
 
 ### Changed
+- **A desktop-sized MCP client budget, optional idle reclamation, and explicit
+  refusal.** The default cap is 64 rather than 16 because desktop clients may
+  keep one stdio bridge per open task; hosts can tune it. Idle expiry is now an
+  opt-in timeout for hosts that know their clients reconnect after EOF. Closing
+  a socket makes `MCPStdioBridge` return immediately even when its stdin owner
+  forgot to close the pipe, while real EOF and transport errors still clean up
+  immediately. At the configured client cap the
+  listener now sends an explicit JSON-RPC capacity error before closing, and a
+  stdio bridge turns that private frame into an actionable stderr message and
+  exit code 2 instead of exiting 0 with empty stdout.
 - **The periodic save writes the cursors that moved, not all of them.** The
   save had one bit of state — "something moved" — and answered it by writing
   every cursor the coordinator held. One harness appending a transcript line a

README.md

@@ -142,6 +142,15 @@ a socket someone is *answering* on is not — the server reports
 `.socketOwnedByAnotherInstance` rather than cutting off the agents attached to
 the other copy.
 
+The default client cap is 64; pass `maximumConnections:` to tune it. Hosts
+whose stdio clients reconnect after EOF may also opt into idle reclamation
+with `idleTimeout:`; it is disabled by default because the library cannot
+assume every client respawns a deliberately closed child. `clientConnections` and
+`onConnectionsChange` expose only a connection id, peer pid, connected time,
+and last activity. A host can call `disconnectClient(id:)` to close one socket
+safely. A client beyond the configured cap receives a framed JSON-RPC
+capacity error instead of a silent close.
+
 `MCPStdioBridge` runs the same binary as a plain stdio MCP server that pumps
 bytes to that socket, which is how MCP clients that spawn a command reach a
 running GUI app:
@@ -163,6 +172,10 @@ if bridge.isRequested() { exit(bridge.run()) }
 The static functions this wraps (`MCPStdioBridge.isRequested`,
 `.socketPath`, `.run`, each taking `config` explicitly) are still there
 unchanged, for callers that would rather not hold an instance.
+If the listener is full, the bridge writes the configured
+`connectionLimitMessage` to stderr and exits with
+`MCPStdioBridge.ExitCode.connectionLimit` (2); it does not report a successful
+empty session.
 
 `MCPJSON`, `MCPRequest`, `MCPResponse`, `MCPTool`, `MCPResource`, and
 `MCPArguments` are here too, so a host writes its tool catalog and its