Files
llmx/codex-rs/app-server/tests/suite/list_resume.rs

309 lines
11 KiB
Rust
Raw Normal View History

use std::fs;
use std::path::Path;
fix: separate `codex mcp` into `codex mcp-server` and `codex app-server` (#4471) This is a very large PR with some non-backwards-compatible changes. Historically, `codex mcp` (or `codex mcp serve`) started a JSON-RPC-ish server that had two overlapping responsibilities: - Running an MCP server, providing some basic tool calls. - Running the app server used to power experiences such as the VS Code extension. This PR aims to separate these into distinct concepts: - `codex mcp-server` for the MCP server - `codex app-server` for the "application server" Note `codex mcp` still exists because it already has its own subcommands for MCP management (`list`, `add`, etc.) The MCP logic continues to live in `codex-rs/mcp-server` whereas the refactored app server logic is in the new `codex-rs/app-server` folder. Note that most of the existing integration tests in `codex-rs/mcp-server/tests/suite` were actually for the app server, so all the tests have been moved with the exception of `codex-rs/mcp-server/tests/suite/mod.rs`. Because this is already a large diff, I tried not to change more than I had to, so `codex-rs/app-server/tests/common/mcp_process.rs` still uses the name `McpProcess` for now, but I will do some mechanical renamings to things like `AppServer` in subsequent PRs. While `mcp-server` and `app-server` share some overlapping functionality (like reading streams of JSONL and dispatching based on message types) and some differences (completely different message types), I ended up doing a bit of copypasta between the two crates, as both have somewhat similar `message_processor.rs` and `outgoing_message.rs` files for now, though I expect them to diverge more in the near future. One material change is that of the initialize handshake for `codex app-server`, as we no longer use the MCP types for that handshake. Instead, we update `codex-rs/protocol/src/mcp_protocol.rs` to add an `Initialize` variant to `ClientRequest`, which takes the `ClientInfo` object we need to update the `USER_AGENT_SUFFIX` in `codex-rs/app-server/src/message_processor.rs`. One other material change is in `codex-rs/app-server/src/codex_message_processor.rs` where I eliminated a use of the `send_event_as_notification()` method I am generally trying to deprecate (because it blindly maps an `EventMsg` into a `JSONNotification`) in favor of `send_server_notification()`, which takes a `ServerNotification`, as that is intended to be a custom enum of all notification types supported by the app server. So to make this update, I had to introduce a new variant of `ServerNotification`, `SessionConfigured`, which is a non-backwards compatible change with the old `codex mcp`, and clients will have to be updated after the next release that contains this PR. Note that `codex-rs/app-server/tests/suite/list_resume.rs` also had to be update to reflect this change. I introduced `codex-rs/utils/json-to-toml/src/lib.rs` as a small utility crate to avoid some of the copying between `mcp-server` and `app-server`.
2025-09-30 00:06:18 -07:00
use app_test_support::McpProcess;
use app_test_support::to_response;
use codex_app_server_protocol::JSONRPCNotification;
use codex_app_server_protocol::JSONRPCResponse;
use codex_app_server_protocol::ListConversationsParams;
use codex_app_server_protocol::ListConversationsResponse;
use codex_app_server_protocol::NewConversationParams; // reused for overrides shape
use codex_app_server_protocol::RequestId;
use codex_app_server_protocol::ResumeConversationParams;
use codex_app_server_protocol::ResumeConversationResponse;
use codex_app_server_protocol::ServerNotification;
use codex_app_server_protocol::SessionConfiguredNotification;
use pretty_assertions::assert_eq;
use serde_json::json;
use tempfile::TempDir;
use tokio::time::timeout;
use uuid::Uuid;
const DEFAULT_READ_TIMEOUT: std::time::Duration = std::time::Duration::from_secs(10);
#[tokio::test(flavor = "multi_thread", worker_threads = 2)]
async fn test_list_and_resume_conversations() {
// Prepare a temporary CODEX_HOME with a few fake rollout files.
let codex_home = TempDir::new().expect("create temp dir");
create_fake_rollout(
codex_home.path(),
"2025-01-02T12-00-00",
"2025-01-02T12:00:00Z",
"Hello A",
Some("openai"),
);
create_fake_rollout(
codex_home.path(),
"2025-01-01T13-00-00",
"2025-01-01T13:00:00Z",
"Hello B",
Some("openai"),
);
create_fake_rollout(
codex_home.path(),
"2025-01-01T12-00-00",
"2025-01-01T12:00:00Z",
"Hello C",
None,
);
let mut mcp = McpProcess::new(codex_home.path())
.await
.expect("spawn mcp process");
timeout(DEFAULT_READ_TIMEOUT, mcp.initialize())
.await
.expect("init timeout")
.expect("init failed");
// Request first page with size 2
let req_id = mcp
.send_list_conversations_request(ListConversationsParams {
page_size: Some(2),
cursor: None,
model_providers: None,
})
.await
.expect("send listConversations");
let resp: JSONRPCResponse = timeout(
DEFAULT_READ_TIMEOUT,
mcp.read_stream_until_response_message(RequestId::Integer(req_id)),
)
.await
.expect("listConversations timeout")
.expect("listConversations resp");
let ListConversationsResponse { items, next_cursor } =
to_response::<ListConversationsResponse>(resp).expect("deserialize response");
assert_eq!(items.len(), 2);
// Newest first; preview text should match
assert_eq!(items[0].preview, "Hello A");
assert_eq!(items[1].preview, "Hello B");
assert_eq!(items[0].model_provider, "openai");
assert_eq!(items[1].model_provider, "openai");
assert!(items[0].path.is_absolute());
assert!(next_cursor.is_some());
// Request the next page using the cursor
let req_id2 = mcp
.send_list_conversations_request(ListConversationsParams {
page_size: Some(2),
cursor: next_cursor,
model_providers: None,
})
.await
.expect("send listConversations page 2");
let resp2: JSONRPCResponse = timeout(
DEFAULT_READ_TIMEOUT,
mcp.read_stream_until_response_message(RequestId::Integer(req_id2)),
)
.await
.expect("listConversations page 2 timeout")
.expect("listConversations page 2 resp");
let ListConversationsResponse {
items: items2,
next_cursor: next2,
..
} = to_response::<ListConversationsResponse>(resp2).expect("deserialize response");
assert_eq!(items2.len(), 1);
assert_eq!(items2[0].preview, "Hello C");
assert_eq!(items2[0].model_provider, "openai");
assert_eq!(next2, None);
// Add a conversation with an explicit non-OpenAI provider for filter tests.
create_fake_rollout(
codex_home.path(),
"2025-01-01T11-30-00",
"2025-01-01T11:30:00Z",
"Hello TP",
Some("test-provider"),
);
// Filtering by model provider should return only matching sessions.
let filter_req_id = mcp
.send_list_conversations_request(ListConversationsParams {
page_size: Some(10),
cursor: None,
model_providers: Some(vec!["test-provider".to_string()]),
})
.await
.expect("send listConversations filtered");
let filter_resp: JSONRPCResponse = timeout(
DEFAULT_READ_TIMEOUT,
mcp.read_stream_until_response_message(RequestId::Integer(filter_req_id)),
)
.await
.expect("listConversations filtered timeout")
.expect("listConversations filtered resp");
let ListConversationsResponse {
items: filtered_items,
next_cursor: filtered_next,
} = to_response::<ListConversationsResponse>(filter_resp).expect("deserialize filtered");
assert_eq!(filtered_items.len(), 1);
assert_eq!(filtered_next, None);
assert_eq!(filtered_items[0].preview, "Hello TP");
assert_eq!(filtered_items[0].model_provider, "test-provider");
// Empty filter should include every session regardless of provider metadata.
let unfiltered_req_id = mcp
.send_list_conversations_request(ListConversationsParams {
page_size: Some(10),
cursor: None,
model_providers: Some(Vec::new()),
})
.await
.expect("send listConversations unfiltered");
let unfiltered_resp: JSONRPCResponse = timeout(
DEFAULT_READ_TIMEOUT,
mcp.read_stream_until_response_message(RequestId::Integer(unfiltered_req_id)),
)
.await
.expect("listConversations unfiltered timeout")
.expect("listConversations unfiltered resp");
let ListConversationsResponse {
items: unfiltered_items,
next_cursor: unfiltered_next,
} = to_response::<ListConversationsResponse>(unfiltered_resp)
.expect("deserialize unfiltered response");
assert_eq!(unfiltered_items.len(), 4);
assert!(unfiltered_next.is_none());
let empty_req_id = mcp
.send_list_conversations_request(ListConversationsParams {
page_size: Some(10),
cursor: None,
model_providers: Some(vec!["other".to_string()]),
})
.await
.expect("send listConversations filtered empty");
let empty_resp: JSONRPCResponse = timeout(
DEFAULT_READ_TIMEOUT,
mcp.read_stream_until_response_message(RequestId::Integer(empty_req_id)),
)
.await
.expect("listConversations filtered empty timeout")
.expect("listConversations filtered empty resp");
let ListConversationsResponse {
items: empty_items,
next_cursor: empty_next,
} = to_response::<ListConversationsResponse>(empty_resp).expect("deserialize filtered empty");
assert!(empty_items.is_empty());
assert!(empty_next.is_none());
// Now resume one of the sessions and expect a SessionConfigured notification and response.
let resume_req_id = mcp
.send_resume_conversation_request(ResumeConversationParams {
path: items[0].path.clone(),
overrides: Some(NewConversationParams {
model: Some("o3".to_string()),
..Default::default()
}),
})
.await
.expect("send resumeConversation");
fix: separate `codex mcp` into `codex mcp-server` and `codex app-server` (#4471) This is a very large PR with some non-backwards-compatible changes. Historically, `codex mcp` (or `codex mcp serve`) started a JSON-RPC-ish server that had two overlapping responsibilities: - Running an MCP server, providing some basic tool calls. - Running the app server used to power experiences such as the VS Code extension. This PR aims to separate these into distinct concepts: - `codex mcp-server` for the MCP server - `codex app-server` for the "application server" Note `codex mcp` still exists because it already has its own subcommands for MCP management (`list`, `add`, etc.) The MCP logic continues to live in `codex-rs/mcp-server` whereas the refactored app server logic is in the new `codex-rs/app-server` folder. Note that most of the existing integration tests in `codex-rs/mcp-server/tests/suite` were actually for the app server, so all the tests have been moved with the exception of `codex-rs/mcp-server/tests/suite/mod.rs`. Because this is already a large diff, I tried not to change more than I had to, so `codex-rs/app-server/tests/common/mcp_process.rs` still uses the name `McpProcess` for now, but I will do some mechanical renamings to things like `AppServer` in subsequent PRs. While `mcp-server` and `app-server` share some overlapping functionality (like reading streams of JSONL and dispatching based on message types) and some differences (completely different message types), I ended up doing a bit of copypasta between the two crates, as both have somewhat similar `message_processor.rs` and `outgoing_message.rs` files for now, though I expect them to diverge more in the near future. One material change is that of the initialize handshake for `codex app-server`, as we no longer use the MCP types for that handshake. Instead, we update `codex-rs/protocol/src/mcp_protocol.rs` to add an `Initialize` variant to `ClientRequest`, which takes the `ClientInfo` object we need to update the `USER_AGENT_SUFFIX` in `codex-rs/app-server/src/message_processor.rs`. One other material change is in `codex-rs/app-server/src/codex_message_processor.rs` where I eliminated a use of the `send_event_as_notification()` method I am generally trying to deprecate (because it blindly maps an `EventMsg` into a `JSONNotification`) in favor of `send_server_notification()`, which takes a `ServerNotification`, as that is intended to be a custom enum of all notification types supported by the app server. So to make this update, I had to introduce a new variant of `ServerNotification`, `SessionConfigured`, which is a non-backwards compatible change with the old `codex mcp`, and clients will have to be updated after the next release that contains this PR. Note that `codex-rs/app-server/tests/suite/list_resume.rs` also had to be update to reflect this change. I introduced `codex-rs/utils/json-to-toml/src/lib.rs` as a small utility crate to avoid some of the copying between `mcp-server` and `app-server`.
2025-09-30 00:06:18 -07:00
// Expect a codex/event notification with msg.type == sessionConfigured
let notification: JSONRPCNotification = timeout(
DEFAULT_READ_TIMEOUT,
fix: separate `codex mcp` into `codex mcp-server` and `codex app-server` (#4471) This is a very large PR with some non-backwards-compatible changes. Historically, `codex mcp` (or `codex mcp serve`) started a JSON-RPC-ish server that had two overlapping responsibilities: - Running an MCP server, providing some basic tool calls. - Running the app server used to power experiences such as the VS Code extension. This PR aims to separate these into distinct concepts: - `codex mcp-server` for the MCP server - `codex app-server` for the "application server" Note `codex mcp` still exists because it already has its own subcommands for MCP management (`list`, `add`, etc.) The MCP logic continues to live in `codex-rs/mcp-server` whereas the refactored app server logic is in the new `codex-rs/app-server` folder. Note that most of the existing integration tests in `codex-rs/mcp-server/tests/suite` were actually for the app server, so all the tests have been moved with the exception of `codex-rs/mcp-server/tests/suite/mod.rs`. Because this is already a large diff, I tried not to change more than I had to, so `codex-rs/app-server/tests/common/mcp_process.rs` still uses the name `McpProcess` for now, but I will do some mechanical renamings to things like `AppServer` in subsequent PRs. While `mcp-server` and `app-server` share some overlapping functionality (like reading streams of JSONL and dispatching based on message types) and some differences (completely different message types), I ended up doing a bit of copypasta between the two crates, as both have somewhat similar `message_processor.rs` and `outgoing_message.rs` files for now, though I expect them to diverge more in the near future. One material change is that of the initialize handshake for `codex app-server`, as we no longer use the MCP types for that handshake. Instead, we update `codex-rs/protocol/src/mcp_protocol.rs` to add an `Initialize` variant to `ClientRequest`, which takes the `ClientInfo` object we need to update the `USER_AGENT_SUFFIX` in `codex-rs/app-server/src/message_processor.rs`. One other material change is in `codex-rs/app-server/src/codex_message_processor.rs` where I eliminated a use of the `send_event_as_notification()` method I am generally trying to deprecate (because it blindly maps an `EventMsg` into a `JSONNotification`) in favor of `send_server_notification()`, which takes a `ServerNotification`, as that is intended to be a custom enum of all notification types supported by the app server. So to make this update, I had to introduce a new variant of `ServerNotification`, `SessionConfigured`, which is a non-backwards compatible change with the old `codex mcp`, and clients will have to be updated after the next release that contains this PR. Note that `codex-rs/app-server/tests/suite/list_resume.rs` also had to be update to reflect this change. I introduced `codex-rs/utils/json-to-toml/src/lib.rs` as a small utility crate to avoid some of the copying between `mcp-server` and `app-server`.
2025-09-30 00:06:18 -07:00
mcp.read_stream_until_notification_message("sessionConfigured"),
)
.await
fix: separate `codex mcp` into `codex mcp-server` and `codex app-server` (#4471) This is a very large PR with some non-backwards-compatible changes. Historically, `codex mcp` (or `codex mcp serve`) started a JSON-RPC-ish server that had two overlapping responsibilities: - Running an MCP server, providing some basic tool calls. - Running the app server used to power experiences such as the VS Code extension. This PR aims to separate these into distinct concepts: - `codex mcp-server` for the MCP server - `codex app-server` for the "application server" Note `codex mcp` still exists because it already has its own subcommands for MCP management (`list`, `add`, etc.) The MCP logic continues to live in `codex-rs/mcp-server` whereas the refactored app server logic is in the new `codex-rs/app-server` folder. Note that most of the existing integration tests in `codex-rs/mcp-server/tests/suite` were actually for the app server, so all the tests have been moved with the exception of `codex-rs/mcp-server/tests/suite/mod.rs`. Because this is already a large diff, I tried not to change more than I had to, so `codex-rs/app-server/tests/common/mcp_process.rs` still uses the name `McpProcess` for now, but I will do some mechanical renamings to things like `AppServer` in subsequent PRs. While `mcp-server` and `app-server` share some overlapping functionality (like reading streams of JSONL and dispatching based on message types) and some differences (completely different message types), I ended up doing a bit of copypasta between the two crates, as both have somewhat similar `message_processor.rs` and `outgoing_message.rs` files for now, though I expect them to diverge more in the near future. One material change is that of the initialize handshake for `codex app-server`, as we no longer use the MCP types for that handshake. Instead, we update `codex-rs/protocol/src/mcp_protocol.rs` to add an `Initialize` variant to `ClientRequest`, which takes the `ClientInfo` object we need to update the `USER_AGENT_SUFFIX` in `codex-rs/app-server/src/message_processor.rs`. One other material change is in `codex-rs/app-server/src/codex_message_processor.rs` where I eliminated a use of the `send_event_as_notification()` method I am generally trying to deprecate (because it blindly maps an `EventMsg` into a `JSONNotification`) in favor of `send_server_notification()`, which takes a `ServerNotification`, as that is intended to be a custom enum of all notification types supported by the app server. So to make this update, I had to introduce a new variant of `ServerNotification`, `SessionConfigured`, which is a non-backwards compatible change with the old `codex mcp`, and clients will have to be updated after the next release that contains this PR. Note that `codex-rs/app-server/tests/suite/list_resume.rs` also had to be update to reflect this change. I introduced `codex-rs/utils/json-to-toml/src/lib.rs` as a small utility crate to avoid some of the copying between `mcp-server` and `app-server`.
2025-09-30 00:06:18 -07:00
.expect("sessionConfigured notification timeout")
.expect("sessionConfigured notification");
let session_configured: ServerNotification = notification
.try_into()
.expect("deserialize sessionConfigured notification");
// Basic shape assertion: ensure event type is sessionConfigured
let ServerNotification::SessionConfigured(SessionConfiguredNotification {
model,
rollout_path,
..
}) = session_configured
else {
unreachable!("expected sessionConfigured notification");
};
assert_eq!(model, "o3");
assert_eq!(items[0].path.clone(), rollout_path);
// Then the response for resumeConversation
let resume_resp: JSONRPCResponse = timeout(
DEFAULT_READ_TIMEOUT,
mcp.read_stream_until_response_message(RequestId::Integer(resume_req_id)),
)
.await
.expect("resumeConversation timeout")
.expect("resumeConversation resp");
let ResumeConversationResponse {
conversation_id, ..
} = to_response::<ResumeConversationResponse>(resume_resp)
.expect("deserialize resumeConversation response");
// conversation id should be a valid UUID
assert!(!conversation_id.to_string().is_empty());
}
fn create_fake_rollout(
codex_home: &Path,
filename_ts: &str,
meta_rfc3339: &str,
preview: &str,
model_provider: Option<&str>,
) {
let uuid = Uuid::new_v4();
// sessions/YYYY/MM/DD/ derived from filename_ts (YYYY-MM-DDThh-mm-ss)
let year = &filename_ts[0..4];
let month = &filename_ts[5..7];
let day = &filename_ts[8..10];
let dir = codex_home.join("sessions").join(year).join(month).join(day);
fs::create_dir_all(&dir).unwrap_or_else(|e| panic!("create sessions dir: {e}"));
let file_path = dir.join(format!("rollout-{filename_ts}-{uuid}.jsonl"));
let mut lines = Vec::new();
// Meta line with timestamp (flattened meta in payload for new schema)
let mut payload = json!({
"id": uuid,
"timestamp": meta_rfc3339,
"cwd": "/",
"originator": "codex",
"cli_version": "0.0.0",
"instructions": null,
});
if let Some(provider) = model_provider {
payload["model_provider"] = json!(provider);
}
lines.push(
json!({
"timestamp": meta_rfc3339,
"type": "session_meta",
"payload": payload
})
.to_string(),
);
// Minimal user message entry as a persisted response item (with envelope timestamp)
lines.push(
json!({
"timestamp": meta_rfc3339,
"type":"response_item",
"payload": {
"type":"message",
"role":"user",
"content":[{"type":"input_text","text": preview}]
}
})
.to_string(),
);
// Add a matching user message event line to satisfy filters
lines.push(
json!({
"timestamp": meta_rfc3339,
"type":"event_msg",
"payload": {
"type":"user_message",
"message": preview,
"kind": "plain"
}
})
.to_string(),
);
fs::write(file_path, lines.join("\n") + "\n")
.unwrap_or_else(|e| panic!("write rollout file: {e}"));
}