Configuration

Fleet is configured through .fleet/config.yml in your repository root. Created by fleet init.

Full reference

.fleet/config.yml

commander:
  model: claude-opus-4-5      # LLM for commander planning (BYOK)
  poll_interval_minutes: 5     # How often commander checks ship progress

execution:
  strategy: spark              # sequential | mapreduce | spark
  stall_threshold_min: 30      # Shadow dispatch after this many minutes of no heartbeat

merge:
  ci_required: true            # Require CI pass before merge queue
  notify: terminal             # terminal | slack

ships:
  - id: ship-a
    adapter: claude            # claude | codex | aider | a2a
    mode: local                # local | remote

Commander settings

Key	Type	Default	Description
model	string	claude-opus-4-5	LLM model for the commander’s planning and coordination
poll_interval_minutes	number	5	How often the commander checks for ship updates

Execution strategies

Strategy	Description
sequential	One mission at a time, in dependency order
mapreduce	All independent missions in parallel, then dependent ones
spark	Parallel DAG + shadow dispatch + fleet brief (recommended)

Merge settings

Key	Type	Default	Description
ci_required	boolean	true	Whether CI must pass before entering merge queue
notify	string	terminal	Where to send merge notifications

Ship configuration

Each entry in ships defines a machine in the fleet.

Key	Type	Description
id	string	Unique identifier for this ship
adapter	string	Which coding agent adapter to use
mode	string	local (same network) or remote (any machine with GitHub access)