What
On your first message, Claude Code waits for MCP servers (add-on tool providers) to start. Fetching the configuration for those servers now gets the full connect timeout, MCP_CONNECT_TIMEOUT_MS, which defaults to 5000 ms. Before, it had a fixed 1000 ms. If the fetch fails outright, Claude Code logs it and carries on without waiting.
Why
Slower configuration sources get more time to respond, and you can adjust that time with MCP_CONNECT_TIMEOUT_MS. A failed fetch no longer holds up your first turn.
Names in the bundleMCP_CONNECT_TIMEOUT_MS
The entry above is what we published on the day. These lines were added later, as Anthropic's own pages caught up, and they sit beside the original rather than replacing it.
Confirmed since
Anthropic's documentation has since written up MCP_CONNECT_TIMEOUT_MS, on Customize sessions in self-hosted environments.
* **Session startup**: before the tool list is first taken, the session waits up to 5 seconds by default for an HTTP or SSE server whose entry sets [`alwaysLoad: true`](/docs/en/mcp#exempt-a-server-from-deferral), or for all servers when y…self-hosted-environments-configuration see the edit
Two sources agreeTwo things we can check say the same as this entry.
Anthropic's documentation agrees
Anthropic's documentation has since written up MCP_CONNECT_TIMEOUT_MS, on Customize sessions in self-hosted environments.
Anthropic's release notes agree
Fixed MCP_CONNECTION_NONBLOCKING=0 giving up on claude.ai connectors after 1s instead of honoring MCP_CONNECT_TIMEOUT_MS