UNPKG

@adonis-agora/agent

Version:

Governed, durable-backed AI agent (chat + tool-calling + governance) for AdonisJS — part of the Agora ecosystem.

45 lines • 1.97 kB
/** * Is this tool part of this deployment right now? `ToolSpec.enabled` and the handler's * `isEnabled()` are ANDed — either one saying no is enough — and both default to yes. * * Resolved per turn rather than at registration, so the answer can come from configuration that * did not exist when the tool was registered. */ export async function isToolEnabled(spec, handler) { const declared = typeof spec.enabled === 'function' ? await spec.enabled() : (spec.enabled ?? true); if (!declared) { return false; } return handler?.isEnabled === undefined ? true : await handler.isEnabled(); } /** * Drop tools this deployment has turned off, before anyone asks who may call them. A disabled tool * is absent, not forbidden — the difference matters, because "forbidden" tells the model (and the * user reading a refusal) that the capability exists. */ export async function filterToolsByEnabled(entries) { const checked = await Promise.all(entries.map(async (entry) => ({ entry, enabled: await isToolEnabled(entry.spec, entry.handler), }))); return checked.filter((row) => row.enabled).map((row) => row.entry); } /** * May this actor use this tool, per the tool's OWN gate? Tools without a `canUse` say yes and are * governed by the `RolesPolicy` alone. */ export async function canActorUseTool(actor, handler) { return handler?.canUse === undefined ? true : await handler.canUse(actor); } /** * Drop tools whose own `canUse` refuses this actor. Runs after the app-wide `RolesPolicy`, and is * additive to it — a tool can narrow who reaches it, never widen. */ export async function filterToolsByCanUse(entries, actor) { const checked = await Promise.all(entries.map(async (entry) => ({ entry, allowed: await canActorUseTool(actor, entry.handler), }))); return checked.filter((row) => row.allowed).map((row) => row.entry); } //# sourceMappingURL=tool-filters.js.map