fix: a minimal fix for json format mismatch in workforce

[cafeii]

Description

A minimal fix for json format mismatch in workforce.

TaskAssignResult requires responses to be in json format, while the old task assignment prompt did not enforce a json format output.

To address this, an additional instruction has been appended to the prompt, instructing the task assigner to provide the task assignment result in json.

TaskAssignResult expects assignee_id to be a str, but by default json.loads converts numeric strings (e.g., "123") into int (e.g., 123). To prevent this, modify the code to use json.loads(parse_int=str).

Fixes #1731

Checklist

Go over all the following points, and put an x in all the boxes that apply.

• I have read the CONTRIBUTION guide (required)
• I have linked this PR to an issue using the Development section on the right sidebar or by adding Fixes #issue-number in the PR description (required)
• I have checked if any dependencies need to be added or updated in pyproject.toml and uv lock
• I have updated the tests accordingly (required for a bug fix or a new feature)
• I have updated the documentation if needed:
• I have added examples if this is a new feature

If you are unsure about any of these, don't hesitate to ask. We are here to help!

[zjrwtx]

thanks @cafeii ,can we add some fail example to show this problem? not sure if there something other problem

[cafeii]

thanks @cafeii ,can we add some fail example to show this problem? not sure if there something other problem

sure, but i'm also not quite certain about the fail examples. I met this bug while solving #1731, which is really lack of information. in fact, all three test scripts in workforce (hackathon_judges.py, multiple_single_agents.py, role_playing_with_agents.py) cannot run successfully due to the json format bug while testing locally.