fix: change codex/sandbox-state/update from a notification to a request (#8142)

Historically, `accept_elicitation_for_prompt_rule()` was flaky because
we were using a notification to update the sandbox followed by a `shell`
tool request that we expected to be subject to the new sandbox config,
but because [rmcp](https://crates.io/crates/rmcp) MCP servers delegate
each incoming message to a new Tokio task, messages are not guaranteed
to be processed in order, so sometimes the `shell` tool call would run
before the notification was processed.

Prior to this PR, we relied on a generous `sleep()` between the
notification and the request to reduce the change of the test flaking
out.

This PR implements a proper fix, which is to use a _request_ instead of
a notification for the sandbox update so that we can wait for the
response to the sandbox request before sending the request to the
`shell` tool call. Previously, `rmcp` did not support custom requests,
but I fixed that in
https://github.com/modelcontextprotocol/rust-sdk/pull/590, which made it
into the `0.12.0` release (see #8288).

This PR updates `shell-tool-mcp` to expect
`"codex/sandbox-state/update"` as a _request_ instead of a notification
and sends the appropriate ack. Note this behavior is tied to our custom
`codex/sandbox-state` capability, which Codex honors as an MCP client,
which is why `core/src/mcp_connection_manager.rs` had to be updated as
part of this PR, as well.

This PR also updates the docs at `shell-tool-mcp/README.md`.

codex-rs/core/src/lib.rs

@@ -35,7 +35,7 @@ pub mod mcp;
 mod mcp_connection_manager;
 pub mod openai_models;
 pub use mcp_connection_manager::MCP_SANDBOX_STATE_CAPABILITY;
-pub use mcp_connection_manager::MCP_SANDBOX_STATE_NOTIFICATION;
+pub use mcp_connection_manager::MCP_SANDBOX_STATE_METHOD;
 pub use mcp_connection_manager::SandboxState;
 mod mcp_tool_call;
 mod message_history;

codex-rs/core/src/mcp_connection_manager.rs

@@ -184,17 +184,20 @@ struct ManagedClient {
 }
 
 impl ManagedClient {
+    /// Returns once the server has ack'd the sandbox state update.
     async fn notify_sandbox_state_change(&self, sandbox_state: &SandboxState) -> Result<()> {
         if !self.server_supports_sandbox_state_capability {
             return Ok(());
         }
 
-        self.client
-            .send_custom_notification(
-                MCP_SANDBOX_STATE_NOTIFICATION,
+        let _response = self
+            .client
+            .send_custom_request(
+                MCP_SANDBOX_STATE_METHOD,
                 Some(serde_json::to_value(sandbox_state)?),
             )
-            .await
+            .await?;
+        Ok(())
     }
 }
 
@@ -253,9 +256,9 @@ impl AsyncManagedClient {
 
 pub const MCP_SANDBOX_STATE_CAPABILITY: &str = "codex/sandbox-state";
 
-/// Custom MCP notification for sandbox state updates.
+/// Custom MCP request to push sandbox state updates.
 /// When used, the `params` field of the notification is [`SandboxState`].
-pub const MCP_SANDBOX_STATE_NOTIFICATION: &str = "codex/sandbox-state/update";
+pub const MCP_SANDBOX_STATE_METHOD: &str = "codex/sandbox-state/update";
 
 #[derive(Debug, Clone, Serialize, Deserialize)]
 #[serde(rename_all = "camelCase")]