Graphorin API reference v0.1.0

---

Graphorin API reference / @graphorin/skills / activation

activation

Activation surface for @graphorin/skills.

Two activation paths are supported per DEC-206:

1. Auto — every step the agent runtime injects the metadata of every registered skill into the system prompt; the model elects a skill by invoking the synthetic activate_skill(name) tool. Skills with disable-model-invocation: true are excluded from the metadata advertisement so the model never sees them.
2. Slash command — the user types /skill:<name> (optionally followed by free-form arguments) and the agent runtime activates the matched skill regardless of disable-model-invocation.

The runtime owns the actual activation; this module only parses the inputs into a structured payload.

Functions

Function	Description
isSlashCommand	Convenience predicate. Returns true when parseSlashCommand would succeed against the supplied body.
parseSlashCommand	Parse a single message body for a /skill:<name> invocation. The grammar accepts: