pub struct CatalogAiSkillCandidate {
pub description: Option<String>,
pub display_name: String,
pub handle: String,
pub handle_expires_at: String,
pub installability: CatalogAiSkillInstallability,
pub kind: CatalogAiSkillCandidateKind,
pub media_type: CatalogAiSkillMediaType,
pub provenance: CatalogAiSkillCandidateProvenance,
pub publisher: Option<String>,
pub source: CatalogCandidateSource,
pub trust: Option<Value>,
}Expand description
An inert AI skill catalog result. Verified Skill candidates may be installable only when the runtime reports installability and the selected session is permitted to plan installation.
Experimental. This type is part of an experimental wire-protocol surface and may change or be removed in future SDK or CLI releases.
Fields§
§description: Option<String>Description taken verbatim from the card. Inert untrusted text.
display_name: StringDisplay name taken verbatim from the card. Inert untrusted text.
handle: StringOpaque, runtime-instance scoped, TTL-bound, single-use handle for this candidate. Carries no readable information and is rejected when stale, replayed, or presented to a different runtime instance. Never logged.
handle_expires_at: StringISO 8601 timestamp after which the handle is stale and will be rejected.
installability: CatalogAiSkillInstallabilityWhether this AI skill candidate can be planned for verified installation in the selected session.
kind: CatalogAiSkillCandidateKindDiscriminator: this candidate describes an AI skill
media_type: CatalogAiSkillMediaTypeMedia type of the underlying AI skill card
provenance: CatalogAiSkillCandidateProvenanceWhere the catalog reference was observed, without the card itself or any content digest.
publisher: Option<String>Publisher taken verbatim from the card. Inert untrusted text.
source: CatalogCandidateSourceWhere the card came from: exactly one of a URL or embedded data, encoded as a tagged union so neither both nor neither can be represented.
trust: Option<Value>Versioned trust metadata observed from the catalog authority. Optional for protocol-3 compatibility with runtimes that predate trust snapshots. A trust-capable runtime emits an explicit snapshot even when the authority omitted or malformed its trust field.