Start a Chat or Simulations test run of the agent's workflow

Pins the exact compiled draft or published version this run tests, so a later draft edit never retroactively changes a run already in progress. Only mode=chat is accepted here, because this is the only run this route can also advance: Simulations are started through POST /v1/agents/{agent_id}/workflow/simulations, Talk via POST /v1/agents/{agent_id}/browser-sessions with workflow_mode=talk, and Phone via POST /v1/agents/{agent_id}/workflow/phone-pin -- each mints its own run through the path that actually drives it. Chat runs are capped per organization to bound concurrent model calls on the platform test credential.

post/v1/agents/{agent_id}/workflow/runs

Request

  • Base URL: https://api.skysay.ai
  • URL: https://api.skysay.ai/v1/agents/{agent_id}/workflow/runs
  • Auth: none declared

Path parameters

agent_idstring required

Request body

mode'chat' required
pinned_source'draft' | 'published'
test_valuesobject

Values for the workflow's declared variables, for this Chat test only, admitted exactly as a Talk test's are: a question that compares an answer with a value compares it with these (none supplied takes its mismatch route). Refused like a call's values: 400 with reason one of call_variables_missing, call_variables_invalid, call_variables_over_budget, workflow_binding_failed, schema_pair_mismatch, per_call_text_on_workflow, agent_not_published, keypad_value_not_dialable, keypad_value_in_text, and field / where naming the variable or line (never a value).

Response

Created test run

Changes

No recorded changes to this operation across all 1 revision of this API.