Loading skill
Install any skill in seconds. Free to start, no credit card required.
Get Started Free →tsinghua-fib-lab/agentsociety-experiment-config
Use when a hypothesis already exists and experiment configuration files need to be created, validated, or revised, especially after the simulation scale budget has been decided.
Passed 14 of 17 test cases with the skill, and 2 of 17 without it.
and 17 more in the full report
15 fixed3 passed either way4 still failing0 broke
Tokens, turns and time describe resource use, not a dollar price.
One of 14 test cases the skill fixed. Same model, same prompt, with and without the skill.
After config_params.py has been written in hypothesis_002/experiment_001/init/, what CLI command executes config_params.py to produce init_config.json and steps.yaml?
| Checked | Without skill | With skill |
|---|---|---|
| Specifies running .agentsociety/bin/ags.py experiment-config run --hypothesis-id 002 --experiment-id 001. | Missed | Passed |
The instructions your model reads when the skill loads.
Generate and validate experiment configuration (init_config.json + steps.yaml) for AgentSociety2 simulations.
SIM_SETTINGS.json exist and need init_config.json + steps.yamlDo NOT use when:
SIM_SETTINGS.json exists yet (use hypothesis skill first)| Action | Command | Purpose |
|---|---|---|
| validate | $PYTHON_PATH .agentsociety/bin/ags.py experiment-config validate --hypothesis-id ID --experiment-id ID | Check setup and module selection |
| prepare | $PYTHON_PATH .agentsociety/bin/ags.py experiment-config prepare --hypothesis-id ID --experiment-id ID | Create init/ directory and template |
| info | $PYTHON_PATH .agentsociety/bin/ags.py experiment-config info --hypothesis-id ID --experiment-id ID | Display selected module details |
| run | $PYTHON_PATH .agentsociety/bin/ags.py experiment-config run --hypothesis-id ID --experiment-id ID | Execute config_params.py to generate files |
| check | $PYTHON_PATH .agentsociety/bin/ags.py experiment-config check --hypothesis-id ID --experiment-id ID | Validate generated config by instantiating modules |
Use the Python interpreter from .env. See CLAUDE.md for setup.
If the experiment size is still open, ask for the simulation scale budget before generating files. Collect the target agent count or range, step budget, runtime budget, and preferred complexity tier. Offer 2-3 options with trade-offs and a recommendation, then use the selected budget to balance agent count, agent complexity, and run length.
HYPOTHESIS.md and SIM_SETTINGS.json already exist for the target experimentscan-modules when neededuser_data/ has been reviewed before generating defaultsuse-dataset or create-datasettest cases passed, vs 2 / 17 without
More skills from the same source and nearby in the registry, with their headline benchmark lift.