fix: retry MCP callTool on any connection error, not just ErrSessionMissing

[dgageot]

When an MCP server restarts, callTool could receive transport errors (EOF, connection reset) instead of ErrSessionMissing depending on timing. Broaden the retry condition to trigger forceReconnectAndWait on any non-context-cancellation error, fixing a flaky TestRemoteReconnectAfterServerRestart.

Assisted-By: docker-agent

[docker-agent]

Assessment: 🔴 CRITICAL

Found 2 issues in the retry logic changes:

• 1 HIGH severity issue
• 1 MEDIUM severity issue

[docker-agent]

🔴 HIGH: Retry logic now retries on ALL errors, not just connection/session errors

The retry condition was broadened from checking specifically for errors.Is(err, mcp.ErrSessionMissing) to checking for any error that is not context.Canceled. This means the code will now trigger reconnection for ALL errors from CallTool, not just session-related errors.

Impact: Application-level errors like authentication failures, invalid tool names, or validation errors will trigger expensive reconnect attempts that won't fix the underlying issue. This wastes resources and adds latency to error responses.

Recommendation: Consider checking for specific connection/transport error types (e.g., net.Error, io.EOF, syscall.ECONNRESET) instead of retrying on all errors. For example:

if err != nil && isConnectionError(err) && ctx.Err() == nil {
    // retry logic
}

func isConnectionError(err error) bool {
    if errors.Is(err, mcp.ErrSessionMissing) {
        return true
    }
    if errors.Is(err, io.EOF) {
        return true
    }
    var netErr net.Error
    if errors.As(err, &netErr) {
        return true
    }
    // Check for syscall errors like ECONNRESET
    return false
}

[docker-agent]

🟡 MEDIUM: Comment claims to retry only connection/session errors but code retries all non-canceled errors

The comment states "If the call failed with a connection or session error (e.g. the server restarted), trigger or wait for a reconnection" but the actual code checks if err != nil && !errors.Is(err, context.Canceled) && ctx.Err() == nil, which retries on ANY error.

Impact: This mismatch between comment and implementation can mislead future maintainers about the actual behavior of the code.

Recommendation: Update the comment to accurately reflect the implementation, or update the implementation to match the comment's intent. For example:

// If the call failed with any error (except context cancellation),
// trigger or wait for a reconnection and retry the call once.

Or implement selective error checking as suggested in the HIGH severity finding above.