Handle required MCP startup failures across components (#10902)

Summary
- add a `required` flag for MCP servers everywhere config/CLI data is
touched so mandatory helpers can be round-tripped
- have `codex exec` and `codex app-server` thread start/resume fail fast
when required MCPs fail to initialize

codex-rs/core/tests/suite/truncation.rs

@@ -426,6 +426,7 @@ async fn mcp_tool_call_output_exceeds_limit_truncated_for_model() -> Result<()>
                     cwd: None,
                 },
                 enabled: true,
+                required: false,
                 disabled_reason: None,
                 startup_timeout_sec: Some(std::time::Duration::from_secs(10)),
                 tool_timeout_sec: None,
@@ -519,6 +520,7 @@ async fn mcp_image_output_preserves_image_and_no_text_summary() -> Result<()> {
                     cwd: None,
                 },
                 enabled: true,
+                required: false,
                 disabled_reason: None,
                 startup_timeout_sec: Some(Duration::from_secs(10)),
                 tool_timeout_sec: None,
@@ -783,6 +785,7 @@ async fn mcp_tool_call_output_not_truncated_with_custom_limit() -> Result<()> {
                     cwd: None,
                 },
                 enabled: true,
+                required: false,
                 disabled_reason: None,
                 startup_timeout_sec: Some(std::time::Duration::from_secs(10)),
                 tool_timeout_sec: None,