Skills
Skills
What a Skill is, how installing works, and what happens on a run.
Availability note — these docs describe both working and planned product areas. Account authentication, administrator-managed AIHK/AICA tenant regions and keys, signed-in Composer runs and public Guard scans are connected. Skills, Memory, Agents, Workflows, Apps, billing, the public SDK/CLI and the unified API are previews unless a section explicitly says otherwise.
Roadmap and package-contract reference: this page describes the intended GenCMS runtime. The current deployment persists workspace records and drafts but does not execute marketplace Skills, index Memory for retrieval, publish packages, or run Agents, Workflows or Apps. See the deployed API reference for the routes available today.
A Skill is a packaged capability: a manifest, an instruction or handler, typed input and output schemas, examples and tests. Installing one adds it to your workspace; running one consumes credits against whichever model executed it.
What makes it a Skill rather than a prompt
- Reusable — installed once, callable from Composer, Agents, Workflows, Apps and the API.
- Inspectable — the manifest, permissions and prompts are readable before you install.
- Versioned — semantic versions with a changelog; business Workflows pin a version.
- Permission-aware — it can only do what it declared, and Guard checks the declaration.
- Testable — sample inputs with expected properties run on every publish.
- Composable — its typed output is another step's typed input.
- Optionally monetizable — free, one-time, subscription or credit-premium.
The run lifecycle
- 1Input is validated against the Skill's input schema.
- 2Entitlement and credit balance are checked.
- 3Any Memory collections you granted are retrieved.
- 4A model is selected according to the Skill's execution mode.
- 5The Skill runs with only its declared permissions.
- 6Output is validated against the output schema; a schema failure is a failed run, not a silent pass.
- 7Usage is recorded — credits, latency, model, schema result.
A failed schema validation still costs credits, because the model ran. Compare a Skill across models before you standardise on one — schema success rates differ more than prose quality does.