Support per-tool MCP output limits (#41421)

## What changed

- Add a positive `output_token_limit` setting to each entry under an MCP server's `tools` configuration.
- Apply the most restrictive limit when plugin and user policies overlap, while keeping approval policy independent.
- Carry the effective MCP output budget in conversation history so tool output, post-tool hook responses, and resumed sessions use the same truncation limit.

## Testing

- Cover configuration parsing, serialization, schema validation, and plugin policy merging.
- Cover MCP output below and above the configured limit, post-tool hook responses, and session resume.

GitOrigin-RevId: d0beb4fca9ba6055d9e1d31c137373b465d50d61

codex-rs/core/src/session/inject.rs

@@ -72,6 +72,7 @@ impl Session {
             && matches!(&item, ResponseItem::Message { role, .. } if role == "developer"))
         .then_some(CodexHarnessMetadata {
             client_authored: true,
+            ..Default::default()
         });
 
         ResponseItemEnvelope { item, metadata }
@@ -82,9 +83,7 @@ impl Session {
         turn_context: &TurnContext,
         items: Vec<ResponseItemEnvelope>,
     ) {
-        if !self.enabled(Feature::RetainClientDeveloperMessages)
-            || items.iter().all(|item| item.metadata.is_none())
-        {
+        if items.iter().all(|item| item.metadata.is_none()) {
             let items = items
                 .into_iter()
                 .map(ResponseItemEnvelope::into_item)