Skip to main content
Open Runs and select an execution. The timeline shows the run trigger, messages, model and tool activity, any human-input request, terminal status, and error details. When a result is weak, review in this order:
  1. Confirm the run received the expected trigger input.
  2. Check that the correct job, agent, credential, and model were used.
  3. Inspect tool arguments and returned data.
  4. Separate missing evidence from poor synthesis.
  5. Update the narrowest layer: connection, job prompt, or agent instructions.
For a completed run, mark whether the result was useful. Vokra records the workspace activation milestone after two scheduled runs are marked useful; this does not change the job or its output. Do not paste secrets into prompts or human-input responses. Provider tokens belong in credential and integration settings.