You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Copy file name to clipboardExpand all lines: README.md
+6-1Lines changed: 6 additions & 1 deletion
Display the source diff
Display the rich diff
Original file line number
Diff line number
Diff line change
@@ -16,6 +16,7 @@ The plugin lets Claude Code launch one Codex agent or several Codex agents in pa
16
16
- Prompt delivery: stdin, not command-line arguments.
17
17
- Codex home: uses the user's Codex home by default; pass `isolated_codex_home: true` to use a temporary Codex home with auth but without inherited `config.toml` MCP servers.
18
18
- Concurrency: Codex processes run through a global queue. Defaults are `CODEX_SUBAGENTS_MAX_GLOBAL_PROCESSES=4` and `CODEX_SUBAGENTS_MAX_PROJECT_PROCESSES=2`.
19
+
- Progress: long-running tools emit MCP `notifications/progress` events when the client supplies a progress token.
19
20
20
21
Optional environment overrides:
21
22
@@ -67,10 +68,12 @@ npm run test:claude-desktop
67
68
68
69
`test:ci` is the GitHub-safe suite. It uses the fake Codex binary and does not require Claude Code, the Codex desktop app, or live model credentials.
69
70
70
-
`test:comprehensive` runs the TypeScript build, unit tests, stdio MCP smoke test, reliability matrix, MCP stress test, Codex desktop runtime probe, Claude plugin validation, and desktop-shipped Claude Code CLI plugin/auth checks. The runtime probe validates local Codex capabilities without invoking a model.
71
+
`test:comprehensive` runs the TypeScript build, unit tests, stdio MCP smoke test, reliability matrix, MCP stress test, MCP progress notification test, Codex desktop runtime probe, Claude plugin validation, and desktop-shipped Claude Code CLI plugin/auth checks. The runtime probe validates local Codex capabilities without invoking a model.
71
72
72
73
`test:stress` uses the fake Codex binary to exercise queued async jobs, noisy output, malformed JSONL, and truncation behavior.
73
74
75
+
`test:progress` verifies that SDK clients receive monotonically increasing MCP progress notifications from blocking, async start, parallel, and wait-style tool calls.
76
+
74
77
`test:claude-orchestration` is an opt-in live Claude Code test. It loads the plugin inside Claude Code, lets Claude call the plugin MCP tools, and uses the fake Codex binary so no Codex model tokens are spent. It is kept out of `test:comprehensive` because it does spend Claude tokens.
75
78
76
79
`test:claude-real-codex` is the full opt-in live path: Claude Code loads the plugin and calls real Codex through the desktop app binary, including one single agent, one parallel run, and one nested Spark subagent run. It spends both Claude and Codex tokens, so it is intentionally not part of the default suite.
Prefer `start_agent_run` or `start_agents_run` for work that may run longer than a normal MCP request. The async job API keeps Claude responsive, supports cancellation, and avoids request failures caused by long-running Codex subprocesses.
111
114
115
+
When a client supports MCP progress tokens, `run_agent`, `run_agents`, `start_agent_run`, `start_agents_run`, `get_agent_run`, `wait_agent_run`, and `cancel_agent_run` send progress notifications. SDK clients should pass an `onprogress` handler and enable timeout reset on progress for long waits.
"test:ci": "npm run build && npm test && npm run smoke:mcp && npm run test:reliability && npm run test:stress",
32
-
"test:comprehensive": "npm run build && npm test && npm run smoke:mcp && npm run test:reliability && npm run test:stress && npm run test:codex-runtime && npm run validate:plugin && npm run test:claude-desktop",
32
+
"test:ci": "npm run build && npm test && npm run smoke:mcp && npm run test:reliability && npm run test:stress && npm run test:progress",
33
+
"test:comprehensive": "npm run build && npm test && npm run smoke:mcp && npm run test:reliability && npm run test:stress && npm run test:progress && npm run test:codex-runtime && npm run validate:plugin && npm run test:claude-desktop",
0 commit comments