Passer au contenu principal
POST
/
agents
/
conversations
/
spans
Spans de conversation GenAI
curl --request POST \
  --url https://api.example.com/agents/conversations/spans \
  --header 'Content-Type: application/json' \
  --data '
{
  "project_id": "<string>",
  "conversation_ids": [
    "<string>"
  ],
  "started_after": "2023-11-07T05:31:56Z",
  "started_before": "2023-11-07T05:31:56Z"
}
'
{
  "conversations": [
    {
      "conversation_id": "<string>",
      "spans": [
        {
          "duration_ms": 123,
          "operation_name": "<string>",
          "span_id": "<string>",
          "trace_id": "<string>"
        }
      ],
      "spans_feedback": [
        {
          "trace_id": "<string>",
          "ratings": [
            {
              "name": "<string>",
              "value": 123,
              "confidence": 123,
              "reason": "<string>"
            }
          ],
          "tags": [
            "<string>"
          ]
        }
      ]
    }
  ]
}

Corps

application/json

Demandez les séquences de spans pour un ensemble explicite de conversations.

Lit uniquement les colonnes scalaires des spans (sans corps de message) pour les conversation_ids donnés. Alimente la mini-carte des spans dans la liste des conversations.

project_id
string
requis
conversation_ids
string[]
Maximum array length: 10000
started_after
string<date-time> | null
started_before
string<date-time> | null

Réponse

Réponse réussie

Séquences de spans + marqueurs de feedback, une entrée par conversation demandée.

conversations
AgentConversationSpans · object[]