Skip to content

The first exam for code agents in the physical world: frozen robot tasks from many benchmark suites, each answered with a program and verified in a separate sandbox. Runs on Harbor.

720 B
Raw Blame
# Harbor forwards these into the agent container. Copy to .env (gitignored) or export them.
ANTHROPIC_API_KEY=            # or CLAUDE_CODE_OAUTH_TOKEN=... with CLAUDE_FORCE_OAUTH=1
OPENAI_API_KEY=               # or CODEX_FORCE_AUTH_JSON=1 to reuse ~/.codex/auth.json (ChatGPT login)
# Through OpenRouter instead: the OpenRouter key as OPENAI_API_KEY with OPENAI_BASE_URL=https://openrouter.ai/api/v1
# (Codex), and as ANTHROPIC_API_KEY with ANTHROPIC_BASE_URL=https://openrouter.ai/api (Claude Code).
# Azure OpenAI: the resource key as OPENAI_API_KEY with OPENAI_BASE_URL=https://<resource>.openai.azure.com/openai/v1
# (or .cognitiveservices.azure.com), and the deployment name as the model (-m openai/<deployment>).