Configuration

reactor init writes a fully-commented reactor.yml at the project root. Every command reads it from <project>/reactor.yml. When the file is absent, the documented defaults apply.

The reactor.yml schema

state:
  dir: ./.reactor              # durable state (receipts, world-models, IR cache)

model:
  provider: openrouter         # openrouter (default) | openai | anthropic | google | <custom>
  render_model: google/gemini-3.5-flash
  compile_model: google/gemini-3.5-flash
  temperature: 0
  max_turns: 200
  # base_url: ...               # optional: override the provider's endpoint
  # api_key_env: ...            # optional: read the key from a different env var

sandbox:
  mode: none                   # none (default) | docker
  shell_timeout_ms: 300000

gateways:                      # external-driven entry points
  - node: inbox
    source_id: inbox
    connector:
      type: static             # static | http | file (or a connectors.{cjs,js} plugin)
      id_field: id
      items: [{ id: item-1, body: "the first item" }]

reactors: []                   # optional: a multi-reactor host (see below)

state

model

The compile model is a component of the content-addressed cache key (contract-set fingerprint, SDK version, model id). Changing it invalidates the cache and forces a recompile.

Choosing a model provider

By default Reactor talks to models through OpenRouter, but it is not bound to it. The CLI drives one bounded @openai/agents session per render/compile step, and any vendor with an OpenAI-compatible Chat Completions endpoint plugs in by naming it in model:. Point at OpenAI, Anthropic, or Google directly -- no OpenRouter account required:

# Anthropic, directly (set ANTHROPIC_API_KEY in your env or a .env)
model:
  provider: anthropic
  render_model: claude-haiku-4-5
  compile_model: claude-haiku-4-5

Each built-in provider resolves to an endpoint + a key env var:

openrouter, openai, and google are OpenAI-compatible Chat Completions surfaces. anthropic is special: the CLI routes it through Anthropic's native Messages API (via the bundled @openai/agents AI-SDK adapter over @ai-sdk/anthropic), because Anthropic's OpenAI-compat endpoint ignores response_format and can't drive Reactor's structured compile/render. You don't install or wire anything -- provider: anthropic just works, structured outputs and all. An optional base_url points the adapter at a proxy in front of the Messages API.

For any other OpenAI-compatible vendor (or a self-hosted gateway), name it freely and supply both base_url and api_key_env:

model:
  provider: together
  base_url: https://api.together.xyz/v1
  api_key_env: TOGETHER_API_KEY
  render_model: meta-llama/Llama-3.3-70B-Instruct-Turbo
  compile_model: meta-llama/Llama-3.3-70B-Instruct-Turbo

api_key_env also overrides a built-in's key var -- handy when, say, your OpenRouter key lives under a different name. Run reactor doctor to confirm the configured provider's key is visible (it reports the exact env var, never printing the value), and reactor doctor --live to drive one real round-trip against it.

sandbox

The sandbox block is the render threat-model knob.

See connectors and sandbox for the full render-isolation behavior, including the Docker-absent fallback.

gateways

Each gateways entry is an external-driven entry point bound to a connector.

See connectors and sandbox for the built-in connector types (static, http, file) and the connectors.{cjs,js} plugin shape.

reactors

A reactors list hosts N isolated reactors in one serve process. When the list is empty (the default), the project is single-reactor and the top-level state and gateways are that one reactor.

Each entry is isolated: its own contracts directory, state directory, substrate, schedule, and cursors, so one reactor never corrupts another. See the multi-reactor host for how --concurrency parallelizes across reactors.

Environment variables

Global flags

These flags are honored by every command and override the file: