Merge branch 'main' into capture-shell-partial-output

codex-rs/app-server/README.md

@@ -90,6 +90,7 @@ Example (from OpenAI's official VSCode extension):
 - `skills/list` — list skills for one or more `cwd` values (optional `forceReload`).
 - `skills/config/write` — write user-level skill config by path.
 - `mcpServer/oauth/login` — start an OAuth login for a configured MCP server; returns an `authorization_url` and later emits `mcpServer/oauthLogin/completed` once the browser flow finishes.
+- `tool/requestUserInput` — prompt the user with 1–3 short questions for a tool call and return their answers (experimental).
 - `config/mcpServer/reload` — reload MCP server config from disk and queue a refresh for loaded threads (applied on each thread's next active turn); returns `{}`. Use this after editing `config.toml` without restarting the server.
 - `mcpServerStatus/list` — enumerate configured MCP servers with their tools, resources, resource templates, and auth status; supports cursor+limit pagination.
 - `feedback/upload` — submit a feedback report (classification + optional reason/logs and conversation_id); returns the tracking thread id.
@@ -470,8 +471,15 @@ Invoke a skill by including `$<skill-name>` in the text input. Add a `skill` inp
   "params": {
     "threadId": "thread-1",
     "input": [
-      { "type": "text", "text": "$skill-creator Add a new skill for triaging flaky CI." },
-      { "type": "skill", "name": "skill-creator", "path": "/Users/me/.codex/skills/skill-creator/SKILL.md" }
+      {
+        "type": "text",
+        "text": "$skill-creator Add a new skill for triaging flaky CI."
+      },
+      {
+        "type": "skill",
+        "name": "skill-creator",
+        "path": "/Users/me/.codex/skills/skill-creator/SKILL.md"
+      }
     ]
   }
 }
@@ -506,10 +514,14 @@ Use `skills/list` to fetch the available skills (optionally scoped by `cwds`, wi
 To enable or disable a skill by path:
 
 ```json
-{ "method": "skills/config/write", "id": 26, "params": {
+{
+  "method": "skills/config/write",
+  "id": 26,
+  "params": {
     "path": "/Users/me/.codex/skills/skill-creator/SKILL.md",
     "enabled": false
-} }
+  }
+}
 ```
 
 ## Auth endpoints

codex-rs/app-server/src/bespoke_event_handling.rs

@@ -54,6 +54,10 @@ use codex_app_server_protocol::ThreadItem;
 use codex_app_server_protocol::ThreadRollbackResponse;
 use codex_app_server_protocol::ThreadTokenUsage;
 use codex_app_server_protocol::ThreadTokenUsageUpdatedNotification;
+use codex_app_server_protocol::ToolRequestUserInputOption;
+use codex_app_server_protocol::ToolRequestUserInputParams;
+use codex_app_server_protocol::ToolRequestUserInputQuestion;
+use codex_app_server_protocol::ToolRequestUserInputResponse;
 use codex_app_server_protocol::Turn;
 use codex_app_server_protocol::TurnCompletedNotification;
 use codex_app_server_protocol::TurnDiffUpdatedNotification;
@@ -83,6 +87,8 @@ use codex_core::review_prompts;
 use codex_protocol::ThreadId;
 use codex_protocol::plan_tool::UpdatePlanArgs;
 use codex_protocol::protocol::ReviewOutputEvent;
+use codex_protocol::request_user_input::RequestUserInputAnswer as CoreRequestUserInputAnswer;
+use codex_protocol::request_user_input::RequestUserInputResponse as CoreRequestUserInputResponse;
 use std::collections::HashMap;
 use std::convert::TryFrom;
 use std::path::PathBuf;
@@ -258,6 +264,57 @@ pub(crate) async fn apply_bespoke_event_handling(
                 });
             }
         },
+        EventMsg::RequestUserInput(request) => {
+            if matches!(api_version, ApiVersion::V2) {
+                let questions = request
+                    .questions
+                    .into_iter()
+                    .map(|question| ToolRequestUserInputQuestion {
+                        id: question.id,
+                        header: question.header,
+                        question: question.question,
+                        options: question.options.map(|options| {
+                            options
+                                .into_iter()
+                                .map(|option| ToolRequestUserInputOption {
+                                    label: option.label,
+                                    description: option.description,
+                                })
+                                .collect()
+                        }),
+                    })
+                    .collect();
+                let params = ToolRequestUserInputParams {
+                    thread_id: conversation_id.to_string(),
+                    turn_id: request.turn_id,
+                    item_id: request.call_id,
+                    questions,
+                };
+                let rx = outgoing
+                    .send_request(ServerRequestPayload::ToolRequestUserInput(params))
+                    .await;
+                tokio::spawn(async move {
+                    on_request_user_input_response(event_turn_id, rx, conversation).await;
+                });
+            } else {
+                error!(
+                    "request_user_input is only supported on api v2 (call_id: {})",
+                    request.call_id
+                );
+                let empty = CoreRequestUserInputResponse {
+                    answers: HashMap::new(),
+                };
+                if let Err(err) = conversation
+                    .submit(Op::UserInputAnswer {
+                        id: event_turn_id,
+                        response: empty,
+                    })
+                    .await
+                {
+                    error!("failed to submit UserInputAnswer: {err}");
+                }
+            }
+        }
         // TODO(celia): properly construct McpToolCall TurnItem in core.
         EventMsg::McpToolCallBegin(begin_event) => {
             let notification = construct_mcp_tool_call_notification(
@@ -1347,6 +1404,66 @@ async fn on_exec_approval_response(
     }
 }
 
+async fn on_request_user_input_response(
+    event_turn_id: String,
+    receiver: oneshot::Receiver<JsonValue>,
+    conversation: Arc<CodexThread>,
+) {
+    let response = receiver.await;
+    let value = match response {
+        Ok(value) => value,
+        Err(err) => {
+            error!("request failed: {err:?}");
+            let empty = CoreRequestUserInputResponse {
+                answers: HashMap::new(),
+            };
+            if let Err(err) = conversation
+                .submit(Op::UserInputAnswer {
+                    id: event_turn_id,
+                    response: empty,
+                })
+                .await
+            {
+                error!("failed to submit UserInputAnswer: {err}");
+            }
+            return;
+        }
+    };
+
+    let response =
+        serde_json::from_value::<ToolRequestUserInputResponse>(value).unwrap_or_else(|err| {
+            error!("failed to deserialize ToolRequestUserInputResponse: {err}");
+            ToolRequestUserInputResponse {
+                answers: HashMap::new(),
+            }
+        });
+    let response = CoreRequestUserInputResponse {
+        answers: response
+            .answers
+            .into_iter()
+            .map(|(id, answer)| {
+                (
+                    id,
+                    CoreRequestUserInputAnswer {
+                        selected: answer.selected,
+                        other: answer.other,
+                    },
+                )
+            })
+            .collect(),
+    };
+
+    if let Err(err) = conversation
+        .submit(Op::UserInputAnswer {
+            id: event_turn_id,
+            response,
+        })
+        .await
+    {
+        error!("failed to submit UserInputAnswer: {err}");
+    }
+}
+
 const REVIEW_FALLBACK_MESSAGE: &str = "Reviewer failed to output a response.";
 
 fn render_review_output_text(output: &ReviewOutputEvent) -> String {

codex-rs/app-server/tests/common/lib.rs

@@ -27,6 +27,7 @@ pub use models_cache::write_models_cache_with_models;
 pub use responses::create_apply_patch_sse_response;
 pub use responses::create_exec_command_sse_response;
 pub use responses::create_final_assistant_message_sse_response;
+pub use responses::create_request_user_input_sse_response;
 pub use responses::create_shell_command_sse_response;
 pub use rollout::create_fake_rollout;
 pub use rollout::create_fake_rollout_with_text_elements;

codex-rs/app-server/tests/common/responses.rs

@@ -60,3 +60,26 @@ pub fn create_exec_command_sse_response(call_id: &str) -> anyhow::Result<String>
         responses::ev_completed("resp-1"),
     ]))
 }
+
+pub fn create_request_user_input_sse_response(call_id: &str) -> anyhow::Result<String> {
+    let tool_call_arguments = serde_json::to_string(&json!({
+        "questions": [{
+            "id": "confirm_path",
+            "header": "Confirm",
+            "question": "Proceed with the plan?",
+            "options": [{
+                "label": "Yes (Recommended)",
+                "description": "Continue the current plan."
+            }, {
+                "label": "No",
+                "description": "Stop and revisit the approach."
+            }]
+        }]
+    }))?;
+
+    Ok(responses::sse(vec![
+        responses::ev_response_created("resp-1"),
+        responses::ev_function_call(call_id, "request_user_input", &tool_call_arguments),
+        responses::ev_completed("resp-1"),
+    ]))
+}

codex-rs/app-server/tests/common/rollout.rs

@@ -59,6 +59,7 @@ pub fn create_fake_rollout(
         cli_version: "0.0.0".to_string(),
         source: SessionSource::Cli,
         model_provider: model_provider.map(str::to_string),
+        base_instructions: None,
     };
     let payload = serde_json::to_value(SessionMetaLine {
         meta,
@@ -136,6 +137,7 @@ pub fn create_fake_rollout_with_text_elements(
         cli_version: "0.0.0".to_string(),
         source: SessionSource::Cli,
         model_provider: model_provider.map(str::to_string),
+        base_instructions: None,
     };
     let payload = serde_json::to_value(SessionMetaLine {
         meta,

codex-rs/app-server/tests/suite/v2/mod.rs

@@ -6,6 +6,7 @@ mod initialize;
 mod model_list;
 mod output_schema;
 mod rate_limits;
+mod request_user_input;
 mod review;
 mod thread_archive;
 mod thread_fork;

codex-rs/app-server/tests/suite/v2/request_user_input.rs

@@ -0,0 +1,127 @@
+use anyhow::Result;
+use app_test_support::McpProcess;
+use app_test_support::create_final_assistant_message_sse_response;
+use app_test_support::create_mock_responses_server_sequence;
+use app_test_support::create_request_user_input_sse_response;
+use app_test_support::to_response;
+use codex_app_server_protocol::JSONRPCResponse;
+use codex_app_server_protocol::RequestId;
+use codex_app_server_protocol::ServerRequest;
+use codex_app_server_protocol::ThreadStartParams;
+use codex_app_server_protocol::ThreadStartResponse;
+use codex_app_server_protocol::TurnStartParams;
+use codex_app_server_protocol::TurnStartResponse;
+use codex_app_server_protocol::UserInput as V2UserInput;
+use codex_protocol::openai_models::ReasoningEffort;
+use tokio::time::timeout;
+
+const DEFAULT_READ_TIMEOUT: std::time::Duration = std::time::Duration::from_secs(10);
+
+#[tokio::test(flavor = "multi_thread", worker_threads = 4)]
+async fn request_user_input_round_trip() -> Result<()> {
+    let codex_home = tempfile::TempDir::new()?;
+    let responses = vec![
+        create_request_user_input_sse_response("call1")?,
+        create_final_assistant_message_sse_response("done")?,
+    ];
+    let server = create_mock_responses_server_sequence(responses).await;
+    create_config_toml(codex_home.path(), &server.uri())?;
+
+    let mut mcp = McpProcess::new(codex_home.path()).await?;
+    timeout(DEFAULT_READ_TIMEOUT, mcp.initialize()).await??;
+
+    let thread_start_id = mcp
+        .send_thread_start_request(ThreadStartParams {
+            model: Some("mock-model".to_string()),
+            ..Default::default()
+        })
+        .await?;
+    let thread_start_resp: JSONRPCResponse = timeout(
+        DEFAULT_READ_TIMEOUT,
+        mcp.read_stream_until_response_message(RequestId::Integer(thread_start_id)),
+    )
+    .await??;
+    let ThreadStartResponse { thread, .. } = to_response(thread_start_resp)?;
+
+    let turn_start_id = mcp
+        .send_turn_start_request(TurnStartParams {
+            thread_id: thread.id.clone(),
+            input: vec![V2UserInput::Text {
+                text: "ask something".to_string(),
+                text_elements: Vec::new(),
+            }],
+            model: Some("mock-model".to_string()),
+            effort: Some(ReasoningEffort::Medium),
+            ..Default::default()
+        })
+        .await?;
+    let turn_start_resp: JSONRPCResponse = timeout(
+        DEFAULT_READ_TIMEOUT,
+        mcp.read_stream_until_response_message(RequestId::Integer(turn_start_id)),
+    )
+    .await??;
+    let TurnStartResponse { turn, .. } = to_response(turn_start_resp)?;
+
+    let server_req = timeout(
+        DEFAULT_READ_TIMEOUT,
+        mcp.read_stream_until_request_message(),
+    )
+    .await??;
+    let ServerRequest::ToolRequestUserInput { request_id, params } = server_req else {
+        panic!("expected ToolRequestUserInput request, got: {server_req:?}");
+    };
+
+    assert_eq!(params.thread_id, thread.id);
+    assert_eq!(params.turn_id, turn.id);
+    assert_eq!(params.item_id, "call1");
+    assert_eq!(params.questions.len(), 1);
+
+    mcp.send_response(
+        request_id,
+        serde_json::json!({
+            "answers": {
+                "confirm_path": { "selected": ["yes"], "other": serde_json::Value::Null }
+            }
+        }),
+    )
+    .await?;
+
+    timeout(
+        DEFAULT_READ_TIMEOUT,
+        mcp.read_stream_until_notification_message("codex/event/task_complete"),
+    )
+    .await??;
+    timeout(
+        DEFAULT_READ_TIMEOUT,
+        mcp.read_stream_until_notification_message("turn/completed"),
+    )
+    .await??;
+
+    Ok(())
+}
+
+fn create_config_toml(codex_home: &std::path::Path, server_uri: &str) -> std::io::Result<()> {
+    let config_toml = codex_home.join("config.toml");
+    std::fs::write(
+        config_toml,
+        format!(
+            r#"
+model = "mock-model"
+approval_policy = "untrusted"
+sandbox_mode = "read-only"
+
+model_provider = "mock_provider"
+
+[features]
+collaboration_modes = true
+
+[model_providers.mock_provider]
+name = "Mock provider for test"
+base_url = "{server_uri}/v1"
+wire_api = "responses"
+request_max_retries = 0
+stream_max_retries = 0
+"#
+        ),
+    )
+}

codex-rs/app-server/tests/suite/v2/turn_start.rs

@@ -1471,8 +1471,18 @@ unified_exec = true
         unreachable!("loop ensures we break on command execution items");
     };
     assert_eq!(completed_id, "uexec-1");
-    assert_eq!(completed_status, CommandExecutionStatus::Completed);
-    assert_eq!(exit_code, Some(0));
+    assert!(
+        matches!(
+            completed_status,
+            CommandExecutionStatus::Completed | CommandExecutionStatus::Failed
+        ),
+        "unexpected command execution status: {completed_status:?}"
+    );
+    if completed_status == CommandExecutionStatus::Completed {
+        assert_eq!(exit_code, Some(0));
+    } else {
+        assert!(exit_code.is_some(), "expected exit_code for failed command");
+    }
     assert_eq!(
         completed_process_id.as_deref(),
         Some(started_process_id.as_str())