Make the HubSpot deals sync fill account_name in core.deals from each deal's associated company on live runs, leaving it empty when there is none or the lookup fails, without making the deal stage events job request companies.
- Version
- Category
- Engineering
- Model configurations
- 2
- Attempts
- 9
- Public trials
- 95 passed
- Median runtime
- 14m 35s
Instruction
# Give HubSpot deals their account names HubSpot rows in `core.deals` have no `account_name`. When the HubSpot deals sync (`backend/jobs/core/deals/hubspot.py`) runs live, fill each deal's `account_name` from the company the deal is associated with in HubSpot. A deal with no associated company has no account name. Two things must not change: - The deal stage events job must not start asking HubSpot for companies. - A company lookup that fails must not fail the sync. Those deals go without an account name for that run. The stack is offline. `python3 /opt/fake-hubspot/server.py` starts a local stand-in for HubSpot's CRM API, with a synthetic portal for each HubSpot credential in the dev database. A job run with `/opt/fake-hubspot/route` on its `PYTHONPATH` reaches the stand-in when it calls `api.hubapi.com`. ## Environment PostgreSQL on localhost:5432 and the local database endpoint the app's driver uses are running when you start, and the dev env files point at the dev database.
Public trials
Results by model
| Model · harness | Mean reward | Range | ± SE | Attempts |
|---|---|---|---|---|
| Claude Opus 5.5Grok Build | 100.0% | 100% – 100% | — | 1 |
| Frontier model ACursor CLI | 50.0% | 0% – 100% | 18.9 pp | 8 |