from copilotkit_intelligence_adk import SkillRegistry, SkillToolset
from google.adk.agents import LlmAgent

registry = SkillRegistry(copilotkit_client, learning_container_id)
agent = LlmAgent(name="skill_agent", tools=[SkillToolset(registry)])

Each tool invocation returns the verified instruction text and its immutable Registry identity metadata. The adapter never executes skill scripts.

Lifecycle and preload

Use await registry.preload() for an explicit fresh preload, await registry.preload_cached() for an explicitly offline preload, or await registry.load() at request time. registry.status, registry.ready, and await registry.wait_until_ready(timeout) expose readiness. A cold registry fails closed until one complete verified set has loaded.

Fresh and cached data

Fresh preload and request-time load delegate only to the generic SDK's client.skills.get(...); cached preload delegates only to client.skills.get_cached(...). Snapshots report fresh or cached source. Request-time loads are throttled for 30 seconds and share one in-flight call. A transient refresh publishes stale and raises its canonical error: there is no implicit stale-data fallback.

Limits and scripts

The complete set is rejected above 128 skills, above 262144 bytes for one SKILL.md, or above 1048576 aggregate bytes. Instructions must be strict UTF-8. Any failure rejects the full set. Manifest files with a script role or a normalized path beginning with scripts/ are denied before content is read.

Telemetry

Events are load.started, load.throttled, load.singleflight_joined, status.changed, load.succeeded, and load.failed. Metadata is restricted to framework/adapter version, source/freshness, status/outcome/reason, latency, skill count, Registry revision, and canonical error code/category/retryability, request ID, and trace ID. Tokens, container IDs, paths, and instruction content are forbidden. A telemetry sink exception is propagated explicitly; every caller joined to that load receives the same adapter failure instance. The adapter version is read from installed distribution metadata, with the source-tree package version as its development fallback. A telemetry callback cannot await a load or close that depends on that callback: a load during load.started and any reentrant close reject immediately, while terminal callbacks observe the already-published ready snapshot or exact failure.

Errors

Auth, permission, HTTP 401/403/404/410, archived-container, project-mismatch, container-not-found, and Registry-unrecoverable errors deny the set and preserve the generic SDK's canonical metadata. Transient or integrity refresh failures become LEARNING_REGISTRY_STALE. Adapter validation errors cover count, individual and aggregate byte limits, strict UTF-8, disabled scripts, and unsupported legacy SDK projections.

Closing

await registry.aclose() is idempotent and changes status to closed. Closing does not cancel an already running native invocation, but every future fresh, cached, or request-time load rejects with LEARNING_REGISTRY_CLOSED.

Compatibility

The exact supported ranges are google-adk>=2.0.0,<3.0.0 and copilotkit>=0.1.95,<1.0.0. The public google.adk.tools.base_toolset.BaseToolset hook and its async get_tools(self, readonly_context=None) method were verified through LlmAgent(..., tools=[toolset]) at the minimum Google ADK 2.0.0 and latest compatible Google ADK 2.5.0.

Ownership and release

The Intelligence/Learning team owns this package. It is independently versioned, tagged, and published, and its release does not require or trigger a release of another adapter.