agent/codex
2
0
Fork 0
mirror of https://github.com/pchuan98/codex.git synced 2026-07-01 00:31:56 +08:00
Code Issues Packages Projects Releases Wiki Activity

[codex] Clarify plugin load and runtime capability stages (#28472)

Browse Source 
## Summary

Plugin loading and auth projection both previously produced
`PluginLoadOutcome`. That made an unfiltered load result look like
runtime-ready capabilities and generated capability summaries before
auth routing had run.

This change keeps loaded plugin records in the cache, applies the
current auth policy in `PluginsManager`, and only then builds
`PluginLoadOutcome` and its summaries. Auth changes still reuse the
cached disk load and re-resolve apps and MCP servers without reloading
plugins.

The updated tests cover cached auth changes and verify that capability
summaries match the effective app/MCP surface.

## Testing

- `just test -p codex-core-plugins`
- `just test -p codex-plugin`
- `just fix -p codex-core-plugins`
This commit is contained in:
xl-openai
2026-06-16 04:57:21 -07:00
committed by GitHub
Unverified
parent ef8eb8bdd9
commit de1f77bfdd
5 changed files with 127 additions and 97 deletions
Download Patch File Download Diff File Expand all files Collapse all files
codex-rs/plugin/src/load_outcome.rs
+3 -1
View File
@@ -83,7 +83,9 @@ pub fn prompt_safe_plugin_description(description: Option<&str>) -> Option<Strin
)
}
/// Outcome of loading configured plugins (skills roots, MCP, apps, errors).
/// Runtime view of loaded plugins and their derived capability summaries.
///
/// Callers must apply any runtime capability policies before constructing this outcome.
#[derive(Debug, Clone, PartialEq)]
pub struct PluginLoadOutcome<M> {
plugins: Vec<LoadedPlugin<M>>,