Python: a small working client

Before you startCopy Anchor Link

Python 3.10 or newer. No extra packages. Complete authentication and set DIMERIQ_BASE_URL and OSPRI_ACCESS_TOKEN as shown in the command-line guide.

Download dimeriq.py and the reviewed panel.json. The script uses HTTPS, a timeout and no automatic write retries.

Read firstCopy Anchor Link

python3 dimeriq.py list-runs
python3 dimeriq.py get-run --run-id YOUR_RUN_ID
python3 dimeriq.py candidates --run-id YOUR_RUN_ID

Expected: JSON with your accessible records. Keep the run's inputs, result scope and evidence when building a report. The candidate example requests the first 20 candidates; the API supports offset and limit for more.

Start approved workCopy Anchor Link

python3 dimeriq.py start-panel --input panel.json

Save the returned id. A timeout does not prove the job failed to start.

Pause and resumeCopy Anchor Link

python3 dimeriq.py pause --run-id YOUR_RUN_ID
python3 dimeriq.py get-run --run-id YOUR_RUN_ID
python3 dimeriq.py resume --run-id YOUR_RUN_ID

Wait for the saved STOPPED checkpoint before resuming. Resume returns a new linked execution ID. Read the lifecycle rules.

Save a decisionCopy Anchor Link

Put the documented job_id, candidate_id, decision, note and decision_id in a reviewed JSON file. Then run:

python3 dimeriq.py save-decision --input decision.json

Reuse a decision ID for a retry of that same decision. Read it back before telling someone it was saved.