Run a single query

Run a query

faraday "Generate a SAR using my interesting_molecules.csv dataset"

You can also point Faraday at a config file:

faraday --config /path/to/faraday.yaml "Generate a SAR using my interesting_molecules.csv dataset"

Common config

Model

llm:
  provider: openai
  model: gpt-5
  api_key_env: OPENAI_API_KEY

Workspace

By default, Faraday uses the current working directory as its workspace. To point it somewhere else, use --workspace-source-root or set app.workspace.source_root in faraday.yaml.

Path note: app.workspace.source_root is a local path in your config. sandbox.workspace.container_path is the path seen inside Docker sandboxes.

app:
  mode: host
  workspace:
    source_root: ./workspace_directory/

sandbox:
  backend: docker
  workspace:
    container_path: /workspace

For the example above, interesting_molecules.csv should be present in ./workspace_directory/.

If you want each run to start from a clean copy of the source files, use:

app:
  workspace:
    source_root: ./workspace_directory/
    init_mode: copy
    copy_root: ./.faraday_runtime/workspace-copies
    keep_copy: false

Outputs

Unless configured otherwise, Faraday writes run outputs to ./run_outputs.

That outputs.root value is typically a local path in your config, even when execution happens in Docker.

outputs:
  root: ./run_outputs

Where outputs go

Each run gets its own directory:

run_<timestamp>_<chat_id>_<query_id>

Inside that directory, Faraday writes:

agent_outputs/
run_artifacts/

So a typical run layout looks like:

./run_outputs/run_<timestamp>_<chat_id>_<query_id>/agent_outputs
./run_outputs/run_<timestamp>_<chat_id>_<query_id>/run_artifacts

agent_outputs/ contains files the agent generated during the run. run_artifacts/ contains structured run metadata such as events, results, and trajectory output.

See Results and artifacts for the full output layout.