mirror of
https://github.com/pchuan98/codex.git
synced 2026-07-01 00:31:56 +08:00
ff50b47dce
## Why Plugin analytics overloaded `plugin_id`: most events used the Codex `<plugin>@<marketplace>` identity, while remote install events used the backend plugin ID. That makes the same field change meaning across event types and complicates downstream identity resolution. This change makes the contract unambiguous: - `plugin_id`: the local Codex `<plugin>@<marketplace>` identity, when resolved - `remote_plugin_id`: the backend plugin identity, when available For a remote install failure that happens before plugin details resolve, `plugin_id` is `null` and `remote_plugin_id` remains populated. ## What changed All six plugin analytics events use the same identity contract: - `codex_plugin_installed` - `codex_plugin_install_failed` - `codex_plugin_uninstalled` - `codex_plugin_enabled` - `codex_plugin_disabled` - `codex_plugin_used` Remote identity is resolved from the current installed-plugin snapshot first, with persisted install metadata as fallback. The telemetry metadata type keeps local identity optional for failures that occur before remote details are available. The app-server test client's manual analytics smokes now find remote mutation events through `remote_plugin_id` and validate that `plugin_id` remains local. ## Remote uninstall Resolve and capture telemetry metadata before removing the local plugin cache, then emit `codex_plugin_uninstalled` after the backend confirms success. The event is also emitted when backend uninstall succeeds but local cache cleanup reports `CacheRemove`. If a concurrent remote-cache refresh removes the local bundle before telemetry capture, the already-fetched remote plugin detail supplies fallback capability metadata. ## Validation - `just test -p codex-analytics` — 82 passed - `just test -p codex-core-plugins` — 271 passed - `just test -p codex-app-server-test-client` — 5 passed - `just test -p codex-plugin` — 3 passed - `just test -p codex-app-server plugin_install` — 37 passed - `just test -p codex-app-server plugin_uninstall` — 10 passed The production app-server install/uninstall flow was also exercised against `plugins~Plugin_f1b845ac33888191ac156169c58733c2` (`build-ios-apps@openai-curated-remote`), and the plugin's original uninstalled state was restored.
149 lines
5.3 KiB
Markdown
149 lines
5.3 KiB
Markdown
# App Server Test Client
|
|
Quickstart for running and hitting `codex app-server`.
|
|
|
|
## Quickstart
|
|
|
|
Run from `<reporoot>/codex-rs`.
|
|
|
|
```bash
|
|
# 1) Build debug codex binary
|
|
cargo build -p codex-cli --bin codex
|
|
|
|
# 2) Start websocket app-server in background
|
|
cargo run -p codex-app-server-test-client -- \
|
|
--codex-bin ./target/debug/codex \
|
|
serve --listen ws://127.0.0.1:4222 --kill
|
|
|
|
# 3) Call app-server (defaults to ws://127.0.0.1:4222)
|
|
cargo run -p codex-app-server-test-client -- model-list
|
|
```
|
|
|
|
`send-message` and `send-message-v2` handle `request_user_input` server requests interactively.
|
|
When Codex asks a question, choose a numbered option (or `o` for a free-form answer when offered)
|
|
and the client will send the response and continue streaming the same turn.
|
|
|
|
## Testing Plugin Analytics
|
|
|
|
The `plugin-analytics-smoke` command exercises `plugin/installed`, plugin
|
|
enable/disable config writes, and a structured plugin mention through one
|
|
app-server connection. Analytics are captured to a local JSONL file and are
|
|
not sent to the analytics backend. The model turn uses a loopback Responses
|
|
API server.
|
|
|
|
The selected plugin must already be installed and enabled remotely, and the
|
|
active Codex profile must be authenticated. On a fresh local cache, the command
|
|
retries ephemeral turns while the installed remote bundle finishes syncing.
|
|
|
|
```bash
|
|
# Build a debug Codex binary; analytics capture is unavailable in release builds.
|
|
cargo build -p codex-cli --bin codex
|
|
|
|
cargo run -p codex-app-server-test-client -- \
|
|
--codex-bin ./target/debug/codex \
|
|
plugin-analytics-smoke \
|
|
--plugin-id linear@openai-curated-remote
|
|
```
|
|
|
|
Use `--capture-file /tmp/plugin-analytics.jsonl` to select the output path.
|
|
The command validates one `codex_plugin_disabled`, `codex_plugin_enabled`, and
|
|
`codex_plugin_used` event with the expected local and remote plugin identities
|
|
and capability metadata. Each event includes the local ID in `plugin_id` and the
|
|
backend ID in `remote_plugin_id`. The enabled and disabled events come from
|
|
successful writes to the temporary config; the command does not mutate the
|
|
remote enabled state. It prints the events and leaves the JSONL file in place
|
|
for inspection. It does not install or uninstall plugins and does not modify
|
|
the profile's persistent config.
|
|
|
|
### Testing remote install and uninstall analytics
|
|
|
|
`plugin-analytics-mutation-smoke` is a manually invoked live smoke test. It
|
|
contacts the configured remote plugin API and temporarily changes the active
|
|
account's installed-plugin state. It is not run by `cargo test`, `just test`,
|
|
or CI.
|
|
|
|
Choose a remote plugin that is available to the active account and is not
|
|
currently installed. The command refuses to run when the plugin is already
|
|
installed, installs it, validates `codex_plugin_installed`, uninstalls it, and
|
|
validates `codex_plugin_uninstalled`, and verifies that the original
|
|
uninstalled state was restored.
|
|
|
|
The mutation events include the local Codex ID in `plugin_id` and the backend ID
|
|
in `remote_plugin_id`.
|
|
|
|
`--remote-plugin-id` takes the backend ID, such as `plugins~Plugin_...`, not the
|
|
local `<plugin>@<marketplace>` ID.
|
|
|
|
```bash
|
|
cargo run -p codex-app-server-test-client -- \
|
|
--codex-bin ./target/debug/codex \
|
|
plugin-analytics-mutation-smoke \
|
|
--remote-plugin-id <REMOTE_PLUGIN_ID> \
|
|
--confirm-account-mutation \
|
|
--capture-file /tmp/plugin-mutation-analytics.jsonl
|
|
```
|
|
|
|
Analytics use the normal queue, reduction, batching, and serialization path,
|
|
but the debug capture destination suppresses analytics network delivery. The
|
|
command prints one of these final states:
|
|
|
|
- `PASS`: the install and uninstall events validated and the plugin is uninstalled.
|
|
- `FAIL-CLEAN`: validation failed, but the original uninstalled state was
|
|
restored.
|
|
- `FAIL-LOCAL-CACHE`: the backend is uninstalled, but local cleanup reported
|
|
an error.
|
|
- `FAIL-DIRTY`: cleanup failed and the plugin still appears installed.
|
|
- `FAIL-UNKNOWN`: the command could not verify the final installed state.
|
|
|
|
For a dirty or uncertain result, retry cleanup with:
|
|
|
|
```bash
|
|
cargo run -p codex-app-server-test-client -- \
|
|
--codex-bin ./target/debug/codex \
|
|
plugin-remote-uninstall \
|
|
--remote-plugin-id <REMOTE_PLUGIN_ID> \
|
|
--confirm-account-mutation
|
|
```
|
|
|
|
Cleanup does not require analytics capture or a debug Codex binary. When the
|
|
smoke uses global `--config` overrides, its printed recovery command preserves
|
|
them so cleanup targets the same backend and account.
|
|
|
|
## Watching Raw Inbound Traffic
|
|
|
|
Initialize a connection, then print every inbound JSON-RPC message until you stop it with
|
|
`Ctrl+C`:
|
|
|
|
```bash
|
|
cargo run -p codex-app-server-test-client -- watch
|
|
```
|
|
|
|
## Testing Thread Rejoin Behavior
|
|
|
|
Build and start an app server using commands above. The app-server log is written to `/tmp/codex-app-server-test-client/app-server.log`
|
|
|
|
### 1) Get a thread id
|
|
|
|
Create at least one thread, then list threads:
|
|
|
|
```bash
|
|
cargo run -p codex-app-server-test-client -- send-message-v2 "seed thread for rejoin test"
|
|
cargo run -p codex-app-server-test-client -- thread-list --limit 5
|
|
```
|
|
|
|
Copy a thread id from the `thread-list` output.
|
|
|
|
### 2) Rejoin while a turn is in progress (two terminals)
|
|
|
|
Terminal A:
|
|
|
|
```bash
|
|
cargo run --bin codex-app-server-test-client -- \
|
|
resume-message-v2 <THREAD_ID> "respond with thorough docs on the rust core"
|
|
```
|
|
|
|
Terminal B (while Terminal A is still streaming):
|
|
|
|
```bash
|
|
cargo run --bin codex-app-server-test-client -- thread-resume <THREAD_ID>
|
|
```
|