# VerifIF Resource Server Configuration
# This server validates LLM responses against instruction-following criteria
# using both fast rule-based validators and async LLM-as-a-judge validators.

verifif:
  resources_servers:
    verifif:
      entrypoint: app.py
      domain: instruction_following
      verified: false
      description: VerifIF instruction following validators with rule-based and LLM judge support
      value: Improve instruction following capabilities with comprehensive validation
      # Reward aggregation: how individual check verdicts combine into the final reward
      # Options: all (AND), any (OR), mean, min, max
      aggregation_mode: all
      # LLM Judge configuration - uses policy model credentials by default
      judge_base_url: ${policy_base_url}
      judge_api_key: ${policy_api_key}
      judge_model: ${policy_model_name}
      # Judge sampling parameters
      judge_temperature: 0.7
      judge_top_p: 0.8
      judge_max_tokens: 10000

verifif_simple_agent:
  responses_api_agents:
    simple_agent:
      entrypoint: app.py
      resources_server:
        type: resources_servers
        name: verifif
      model_server:
        type: responses_api_models
        name: policy_model
      datasets:
      - name: example
        type: example
        jsonl_fpath: resources_servers/verifif/data/example.jsonl