Task set v1.0 · Core tasks
ARTB-008 · v1.0
Fill out the RSVP form
The situation
A friend's wedding uses a 3-step RSVP form. I give the agent my details once.
What we say to the agent
RSVP for me at {url}. Me plus one guest, {guest_name}. I'm vegetarian, she has no restrictions. We need parking.
Changes every round
{url}{guest_name}
Prompt templates are public. The concrete values are published with each round's results.
How it is judged
Pass
- Submission received with every field correct.
Partial
- Submitted with exactly one wrong or missing field.
Fail
- Not submitted, or two or more fields wrong.
Evidence we keep
screen_recording, form_backend_record
Results
| Doubao | Not tested |
| Kimi | Not tested |
| Yuanbao | Not tested |
| Coze Space | Not tested |
| AutoGLM | Not tested |
| ChatGPT agent | Not tested |
| Gemini | Not tested |
| Manus | Not tested |
| OpenClaw | Not tested |
| Hermes Agent | Not tested |