Record attempted tool calls in response metadata (#36181)

## What changed

- Add the opt-in `executed_tool_call_metadata` feature to record model-attempted direct and code-mode tool calls.
- Attach recorded names and arguments to the matching tool output in the next Responses request, preserving the metadata across sampling retries.
- Bound pending calls and argument sizes, and emit truncation metadata when limits are exceeded.

## Testing

- Cover disabled-by-default behavior, namespaced and nested calls, blocked or failed attempts, retry handling, and recorder limits.

GitOrigin-RevId: 5ce917b73e797b0f6904d46477b9d2cf8ea71bf2
This commit is contained in:
ningyi-oai
2026-07-30 16:22:43 +00:00
committed by copyberry
parent 355d2a802a
commit ceb4bc72c4
16 changed files with 570 additions and 0 deletions
+5
View File
@@ -1082,6 +1082,10 @@ impl Session {
}).await;
}
let executed_tool_calls = config
.features
.enabled(Feature::ExecutedToolCallMetadata)
.then(|| Arc::new(crate::state::ExecutedToolCallRecorder::default()));
let services = SessionServices {
// Start with an empty connection set. The initialized set is
// published after SessionConfigured so MCP events follow it.
@@ -1159,6 +1163,7 @@ impl Session {
session_configuration.parent_thread_id,
),
),
executed_tool_calls,
code_mode_service: crate::tools::code_mode::CodeModeService::new(
Arc::clone(&code_mode_session_provider),
&config.features,
+10
View File
@@ -5718,6 +5718,10 @@ pub(crate) async fn make_session_and_context() -> (Session, TurnContext) {
));
let network_approval = Arc::new(NetworkApprovalService::default());
let mcp_runtime = Arc::new(codex_mcp::McpRuntime::empty(config.prefix_mcp_tool_names()));
let executed_tool_calls = config
.features
.enabled(Feature::ExecutedToolCallMetadata)
.then(|| Arc::new(crate::state::ExecutedToolCallRecorder::default()));
let services = SessionServices {
mcp_runtime,
unified_exec_manager: UnifiedExecProcessManager::new(
@@ -5793,6 +5797,7 @@ pub(crate) async fn make_session_and_context() -> (Session, TurnContext) {
/*attestation_provider*/ None,
config.http_client_factory(),
),
executed_tool_calls,
code_mode_service: crate::tools::code_mode::CodeModeService::new(
Arc::new(codex_code_mode::InProcessCodeModeSessionProvider),
&config.features,
@@ -7901,6 +7906,10 @@ where
));
let network_approval = Arc::new(NetworkApprovalService::default());
let mcp_runtime = Arc::new(codex_mcp::McpRuntime::empty(config.prefix_mcp_tool_names()));
let executed_tool_calls = config
.features
.enabled(Feature::ExecutedToolCallMetadata)
.then(|| Arc::new(crate::state::ExecutedToolCallRecorder::default()));
let services = SessionServices {
mcp_runtime,
unified_exec_manager: UnifiedExecProcessManager::new(
@@ -7976,6 +7985,7 @@ where
/*attestation_provider*/ None,
config.http_client_factory(),
),
executed_tool_calls,
code_mode_service: crate::tools::code_mode::CodeModeService::new(
Arc::new(codex_code_mode::InProcessCodeModeSessionProvider),
&config.features,
+8
View File
@@ -1336,6 +1336,7 @@ async fn run_sampling_request(
let mut retries = 0;
let mut initial_input = Some(input);
let mut original_input = None;
let mut executed_tool_calls_by_output = HashMap::new();
loop {
let prompt_input = if let Some(input) = initial_input.take() {
input
@@ -1344,6 +1345,13 @@ async fn run_sampling_request(
.await
.for_prompt(&turn_context.model_info.input_modalities)
};
let mut prompt_input = prompt_input;
if let Some(executed_tool_calls) = sess.services.executed_tool_calls.as_ref()
&& executed_tool_calls
.attach_pending_to_prompt(&mut prompt_input, &mut executed_tool_calls_by_output)
{
codex_protocol::models::bound_executed_tool_calls_for_prompt(&mut prompt_input);
}
let prompt = build_prompt(
prompt_input,
router.as_ref(),