## Why
Agent Plugin manifests previously skipped app declarations, including service-generated declarations installed with remote curated plugins. This prevented those plugins from using app routing and suppressing matching direct MCP connections.
## What changed
- Recognize app declarations for Agent Plugins from `openai-curated-remote` across loading, capability summaries, plugin details, tool suggestions, and install authentication.
- Apply app routing to suppress same-name MCP servers when an app route is available. Keep unrelated MCP servers available and preserve direct MCP access in API-key mode.
- Keep app declarations inactive for Agent Plugins from other marketplaces.
## Testing
Add coverage for marketplace and authentication routing, service-generated app declarations during installation with both manifest formats, and app tools reaching turns. Verify that disabling Apps does not restore matching direct MCP connections.
GitOrigin-RevId: 791739569f0e01c898040bc2d028cf517f767e00
## Why
Repeated plugin discovery and loading reparse unchanged manifests and repeat warnings for invalid fields.
## What changed
- Share a manifest cache through `PluginStore` across marketplace listing, plugin loading, capability inspection, and tool suggestion metadata.
- Reuse successful parses by plugin root, manifest path, and content digests, including the Codex overlay. Reread files on each load so manifest edits and overlay changes or removal remain visible.
- Bound the cache to 128 entries and bypass caching for inputs larger than 64 KiB. Retry failed parses and share entries across store clones while keeping independent stores isolated.
- Reuse loaded manifest data when loading apps and MCP metadata.
## Testing
Add unit coverage for parse reuse, warning deduplication, revision invalidation, repair after parse failures, equivalent Windows paths, store isolation, and oversized inputs. Add an app-server regression test that verifies repeated `plugin/list` requests observe manifest and overlay revisions in one running server.
GitOrigin-RevId: ed082cd7d198015cd8ed35db7f2af74e13f49915
## What changed
- Define skill config selectors, ordered rules, and layer-stack parsing in `codex-config`.
- Resolve disabled skill paths from generic skill name/path pairs, so configuration logic no longer depends on `SkillMetadata`.
- Update plugin and host skill consumers to use the `codex-config` API and consolidate the rule tests with the skill config tests.
GitOrigin-RevId: 3fe67869708df2652befe28d58cdeba933256f84
## What changed
- Use `HostSkillRoot` throughout host root resolution and loading, including plugin roots.
- Remove the duplicate loader, root snapshot, and product-filtering implementation from `core-skills` now that loading is owned by the skills extension.
- Move and expand loader coverage for discovery, namespaces, filesystem routing, root merging, symlinks, and frontmatter parsing.
## Testing
- Added focused unit and integration tests under `ext/skills` and `skills` for the consolidated loader behavior.
GitOrigin-RevId: 214d06d59bf3033ee0f220ea5959ee4feff66782
## Why
Plugin suggestions should report skills only when the plugin has an enabled skill.
## What changed
- Load suggestion metadata through the shared plugin skill-root loader.
- Apply skill configuration when setting `has_skills` on suggestion candidates.
## Testing
Added coverage for local plugin suggestions with the plugin skill both enabled and disabled.
GitOrigin-RevId: bceb14528c9042b474eaad6efaf919bef72e0ca4
## What changed
- Track Agent Plugin manifests through plugin, skill, and MCP loading so their capabilities use format-specific behavior without changing legacy plugins.
- Discover only direct-child skills, exclude app and hook capabilities, isolate MCP data, and reject MCP configuration files that are non-regular or resolve outside the plugin root.
- Bound model-visible skill instructions, plugin instructions, MCP descriptions, schemas, individual tools, and the aggregate Agent Plugin MCP tool set.
- Stop MCP and OAuth redirects when Agent Plugins send configured or authorization headers, while retaining existing redirect behavior for legacy MCP servers.
## Testing
- Add coverage for capability filtering, skill discovery, isolated MCP data and reserved-path expansion, unsafe MCP configuration files, context limits, and redirect handling.
GitOrigin-RevId: c9af66b051269f3226628ca280a58d32c808c38f
## What changed
- Carry a plugin's local and remote identities together from plugin loading into
`SkillMetadata`.
- Resolve remote IDs from the installed-plugin snapshot when available, falling
back to persisted install metadata only when no snapshot exists.
- Include plugin identity in skill cache keys so identity changes refresh cached
skill metadata.
## Testing
- Cover snapshot and persisted identity resolution, local marketplace isolation,
cached skill refreshes, and propagation through plugin skill snapshots.
GitOrigin-RevId: aabeeb631a43361fe817358ace7f1ea8ba5db708
## What changed
- Define and export skill metadata, policy, dependency, interface, and configuration rule types from `codex-skills`.
- Update core, plugin, and extension consumers to use the shared models while retaining compatibility re-exports from `codex-core-skills`.
- Share product-restriction handling between host and environment skill metadata.
## Testing
- Add unit coverage for default and explicit implicit-invocation and product policies.
GitOrigin-RevId: 9093469f8fff8e2d902dc9781235f64709827a87
## What changed
- Scan up to eight skill roots concurrently instead of loading them one at a time.
- Share scan capacity across host-skill and plugin-skill loads to bound aggregate filesystem work.
- Merge completed scans in input order so root precedence and error ordering remain deterministic.
- Skip canonicalization and parsing work for empty roots.
## Testing
- Add coverage for out-of-order scan completion and shared-capacity blocking.
GitOrigin-RevId: 33a47aaa02aee7fe31a050bb04d51c1e2043d6ae
## Why
Authentication mode is a domain concept used by login, model selection,
telemetry, and transports. Keeping the canonical type in app-server
protocol forces those lower-level crates to depend on an unrelated wire
API.
## What changed
- Added canonical `codex_protocol::auth::AuthMode` domain values.
- Kept the app-server wire DTO unchanged and added an explicit app-side
conversion.
- Removed production app-server-protocol dependencies from login,
model-provider-info, models-manager, and otel call paths.
## Stack
This is PR 2 of 6, stacked on [PR
#29714](https://github.com/openai/codex/pull/29714). Review only the
delta from `codex/split-json-rpc-protocols`. Next: [PR
#29722](https://github.com/openai/codex/pull/29722).
## Validation
- Auth and login coverage passed in the focused protocol/domain test
run.
- App-server account and auth conversion coverage passed.
## Summary
- Preserve raw plugin skill-root snapshots in the matching loaded-plugin
cache entry, keyed by the effective plugin root identity including
namespace.
- Pass those snapshots through `SkillsLoadInput` as an optional preload,
so session startup reuses plugin parsing while ordinary skill loads pass
`None`.
- Keep plugin skill loading cohesive: the existing loaders accept the
optional snapshots directly, and uncached or marketplace-detail paths do
not create a cache.
## Why
Plugin discovery already parses plugin skills to determine available
capabilities. Cold session startup then scanned and parsed the same
roots again while building the skills snapshot.
This solves the same duplicate-work problem as #28623 while keeping
ownership narrow: `PluginsManager` creates and owns
`PluginSkillSnapshots` only for its loaded-plugin cache entry;
`SkillsService` consumes an optional clone. Entry replacement or
clearing naturally drops the snapshots, with no separate generation,
capacity policy, or watcher coupling.
## Validation
- `cargo clippy -p codex-core-skills --all-targets -- -D warnings`
- `just test -p codex-core-plugins
skills_service_reuses_skills_parsed_during_plugin_load`
- `just test -p codex-core-skills
namespaces_plugin_skills_using_provided_namespace`
- `just fmt`
## Why
`built_tools` runs for every sampling request, and local plugin
discovery was repeatedly rereading plugin manifests, skills, MCP
configuration, and app declarations to build the same tool-suggest
metadata.
That source-derived metadata is stable until the existing plugin manager
reloads its cache. Runtime eligibility still needs to reflect the
current install, disable, policy, app-overlap, and authentication state.
## What changed
- Add a bounded, in-memory tool-suggest metadata cache owned by
`PluginsManager`.
- Key cached metadata by plugin identity and source, while applying
authentication routing each time the metadata is projected.
- Invalidate the metadata alongside the existing loaded-plugin cache,
including its normal configuration, marketplace refresh, and
remote-installed-plugin invalidation paths.
- Guard against an in-flight load repopulating stale metadata after
invalidation.
- Keep marketplace membership and all runtime eligibility filtering live
rather than introducing a separate catalog or revision model.
## Impact
Repeated sampling requests reuse already-loaded plugin capability
metadata while retaining the existing plugin-manager lifecycle as the
single freshness boundary.
## Validation
- `just test -p codex-core-plugins` — 252 passed
- Added focused coverage for cache invalidation and authentication
reprojection.
thomas