Files
Assistant/contracts/capability/multimodal.go
T
m 78e7e3fedc
CI / test (push) Successful in 1m9s
CI / test (pull_request) Successful in 1m7s
Add local speech and vision chat skills
2026-08-25 12:26:41 +03:00

21 lines
801 B
Go

package capability
// MultimodalOperation identifies one bounded local media skill. It is kept
// separate from general cognition so health, routing and failure evidence do
// not collapse into a generic model response.
type MultimodalOperation string
const (
MultimodalTranscribe MultimodalOperation = "TRANSCRIBE_AUDIO"
MultimodalUnderstand MultimodalOperation = "UNDERSTAND_IMAGE"
MultimodalRecognizeOCR MultimodalOperation = "RECOGNIZE_OCR"
)
type MultimodalResponse struct {
Operation MultimodalOperation `json:"operation"`
Text string `json:"text"`
Provider string `json:"provider,omitempty"`
ObservedAt int64 `json:"observed_at_unix_millis"`
ConfidenceNote string `json:"confidence_note,omitempty"`
}