TRtrajectory
diagnostics

Ingest Task For Diagnostics

Changed on

Register one task as a private benchmark and return its ingestion result. Call buildBenchmarkImages, poll every page of listBenchmarkImages until each image is ready, then call startTaskDiagnostics with the returned bench_id.

post/api/v1/diagnostics/task/ingest

Request

  • Base URL
    https://api.trajectory.ai (the document declares no server; this is the origin it was published from)
  • URL
    https://api.trajectory.ai/api/v1/diagnostics/task/ingest
  • Auth
    none declared

Query parameters

agent_idstring required
bench_idstring nullable

Request body

namestring required
idstring nullable
split'train' | 'test'
reference_trajectory_idstring nullable
imageunknown
run_commandstring required

For SDK evaluation and training, the command to run your agent in the sandbox, e.g. python /app/agent.py; include the executable, not only its arguments. Your script must call the model, run any tools, log a reward, and finalize the trajectory before exiting.

env_varsobject

Keys must match [A-Za-z_][A-Za-z0-9_]* and cannot start with TRAJECTORY_ or use MODEL_ENDPOINT_ID, MODEL_ENDPOINT_ACCESS_TOKEN, or MODEL_ENDPOINT_URL.

specobject
task_spec_uristring nullable
tagsstring[]

Up to 20 distinct, non-empty task tags of at most 64 characters each.

Response

Successful Response

bench_idstring required
namestring required
task_countinteger required
warningsstring[]

Provider resource adjustments made before ingestion.

Changes