[codex] Remove hardcoded app ID filters#28947
Conversation
There was a problem hiding this comment.
💡 Codex Review
Here are some automated review suggestions for this pull request.
Reviewed commit: 64d94c5571
ℹ️ About Codex in GitHub
Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you
- Open a pull request for review
- Mark a draft as ready
- Comment "@codex review".
If Codex has suggestions, it will comment; otherwise it will react with 👍.
Codex can also answer questions or update the PR. Try commenting "@codex address that feedback".
| if server_name == CODEX_APPS_MCP_SERVER_NAME { | ||
| return Ok(filter_disallowed_codex_apps_tools(tools)); | ||
| } | ||
| Ok(tools) |
There was a problem hiding this comment.
Add integration coverage for agent-visible app tools
This removal changes which Codex Apps MCP tools can reach an agent turn, but the diff only updates codex-mcp cache coverage and app-server list coverage; without a core/suite test that drives a turn with one of the formerly filtered connector IDs, the model-visible behavior can regress while these tests still pass. The repository requires integration tests for agent-logic changes.
AGENTS.md reference: AGENTS.md:L114-L118
Useful? React with 👍 / 👎.
Summary
codex-logindependency fromcodex-utils-pluginsWhy
The client-side policy was duplicated across crates, used opaque IDs without ownership or expiry information, and could drift between app listing and MCP tool behavior. Server-provided visibility, authorization, plugin discoverability, accessibility, enabled-state handling, and consequential-tool approval templates remain unchanged.
Validation
just fmtjust bazel-lock-updatejust bazel-lock-checkgit diff --checkA targeted
codex-mcptest build spent an unusually long time in local compilation/linking. Its first attempt exposed a test-onlyPartialEqassertion issue, which was corrected. A follow-up non-linkingcargo check -p codex-mcp --testswas still running when this draft was opened; CI should provide the complete Rust validation.