@suluk/core - v0.1.13
    Preparing search index...

    Interface SulukResource

    A loadable RESOURCE (C036) — a member of the top-level x-suluk-resources catalog. Cloudflare's "Agent Skills" model, in Suluk's contract-first form: an on-demand bundle of content the agent loads only when a task matches, so a large library does not bloat every prompt. NOT a SulukSkillRef (which is model-bearing, always-on system text) — a resource carries NO model and is content-only. Structural; never read by the matcher.

    interface SulukResource {
        description: string;
        kind: "instructions" | "reference" | "script";
        provenance: { source: string; contentHash: string; version?: string };
        trust?: "author-declared" | "retrieved";
        [ext: `x-${string}`]: unknown;
    }

    Indexable

    • [ext: `x-${string}`]: unknown
    Index

    Properties

    description kind provenance trust?

    Properties

    description: string

    required, routing-oriented — the catalog-listing text the model sees and selects on (mirrors a CF skill description).

    kind: "instructions" | "reference" | "script"

    the kind of loadable content: instructions (a SKILL.md the agent activates), reference (a bundled doc it reads), script (a bundled script it can run — CF script execution is EARLY/experimental).

    provenance: { source: string; contentHash: string; version?: string }

    single source of truth + staleness binding — the catalog/SKILL.md is GENERATED from source, hashed to detect drift (the same pointer-not-prose discipline as a skill, C027).

    trust?: "author-declared" | "retrieved"

    author-declared (trusted) vs retrieved (untrusted) — a retrieved resource may NOT escalate scope/provenance (mirrors SulukSkillRef.trust).

    Settings

    Member Visibility

    On This Page

    Properties