faradayai 0.1.0

Faraday research agent

Faraday Agent

Accelerating science with agentic AI for research and drug design

Faraday is an AI scientist that carries out research workflows end to end:

Install

pip install faradayai

Requires Python 3.10+. For a development install from the repo root:

pip install -e .

---

Credentials

Set a model API key before running. Faraday supports OpenAI, OpenRouter, and Azure OpenAI.

OpenAI (default)

export OPENAI_API_KEY=...

OpenRouter

export OPENROUTER_API_KEY=...

# faraday.yaml
llm:
  provider: openrouter
  model: openai/gpt-5
  api_key_env: OPENROUTER_API_KEY

Azure OpenAI

# faraday.yaml
llm:
  provider: azure
  model: gpt-5
  api_key: OPENAI_API_KEY
  base_url: AZURE_OPENAI_BASE_URL
  api_version: preview

Run faraday --check-tools to verify keys and tool availability after setup.

---

Quickstart

Run a one-shot task:

faraday "Summarize structure–activity relationships across KRAS inhibitors, focusing on motifs that improve binding affinity and selectivity"

Interactive mode:

faraday

With a config file:

faraday --config faraday.example.yaml "Your task here"

Results are written to ./run_outputs/ by default:

run_outputs/
  run_{timestamp}_{chat_id}_{query_id}/
    agent_outputs/       # files, plots, and generated artifacts
    run_artifacts/
      events.jsonl       # timestamped stream of every agent event
      metadata.json      # run parameters (model, steps, config)
      result.json        # final answer and run summary
      trajectory.json    # full trajectory for replay

Useful flags: --model, --max-steps, --debug, --check-tools, --app-mode, --sandbox-backend, --batch-file, --use-docker. See faraday --help for the full list.