init, doctor, status

init

Run brigade init --target <dir> for the interactive picker:

brigade init --target ~/agent-kitchen

A non-interactive shell needs the flags instead: bare brigade init errors with brigade init needs a TTY for the interactive prompt. Pass --depth and --harnesses for scripting.

For CI or scripts, pass flags directly:

brigade init --target ~/agent-kitchen --depth workspace --harnesses claude,codex,openclaw
brigade init --target ./repo --depth repo --harnesses codex
brigade init --target ./repo --harnesses none           # generic install

Add --dry-run to preview the full file-by-file list without writing anything. Files that already exist are marked as refused-without-force rather than implying they would be overwritten.

The workspace config directory is .brigade (older .solo-mise installs are still read), and the solo-mise command is a deprecated alias for brigade.

brigade --help opens with a short start-here block and lists commands in five groups: the core memory loop, the daily operator loop, stations and tools, review/security/research, and wiring and advanced. Every command stays listed; the grouping just replaces the flat dump.

doctor

brigade doctor verifies the wiring. It checks installed bootstrap files against hard byte budgets so overgrown files fail before agents load truncated context. It folds installed managed tools into its report and surfaces each tool’s own health. A missing optional tool is not a failure; missing optional managed tools collapse into one non-failing summary line telling you to run brigade add <station>, instead of a per-tool wall.

brigade doctor

Add --json to emit the machine-readable report instead of text, --target/-t <dir> to point at another install, and --harness {generic,openclaw,hermes} to run a specific harness’s checks.

A green doctor looks like this (abbreviated; a fresh generic install reports dozens of checks):

brigade doctor: target /home/you/agent-kitchen
  harnesses: (none) (owner=this-repo, depth=workspace)
  [ok]   bootstrap: AGENTS.md                /home/you/agent-kitchen/AGENTS.md
  [ok]   bootstrap: MEMORY.md                /home/you/agent-kitchen/MEMORY.md
  [ok]   bootstrap: TOOLS.md                 /home/you/agent-kitchen/TOOLS.md
  [ok]   bootstrap-budget: AGENTS.md         11615/12000 bytes
  [ok]   agents-quality: AGENTS.md           states a definition of done and a handoff path
  [ok]   memory: cards/                      /home/you/agent-kitchen/memory/cards (11 cards)
  [ok]   handoff-source: handoff_sources     not configured; no pending handoffs
  [ok]   publish: hooks/pre-push             /home/you/agent-kitchen/hooks/pre-push
  [ok]   content guard                       embedded scanner available, public-repo policy loads

summary: 53 checks, 0 failed, 0 manual

Anything [warn] is fine; [fail] means the install is incomplete. The openclaw and hermes harnesses add their own checks on top.

status

Once installed, brigade status reports over the station registry as a per-station table.

brigade status

Add --json to emit the machine-readable report instead of text, and --target/-t <dir> to point at another install.

brigade status: /home/you/agent-kitchen
  core           [ok]  16 ok, 1 warn, 0 fail  - workspace bootstrap and harness adapters
  memory         [ok]  15 ok, 2 warn, 0 fail  - handoff inbox, ingest, and memory-care
  guard          [ok]  2 ok, 0 warn, 0 fail  - publish safety and content scrub
  tokens         [empty]  0 ok, 0 warn, 0 fail  - output compaction
  search         [empty]  0 ok, 0 warn, 0 fail  - local semantic code search
  security       [ok]  2 ok, 3 warn, 0 fail  - agent workspace security scanning
  pantry         [empty]  0 ok, 0 warn, 0 fail  - agent session auth sync
  notifications  [empty]  0 ok, 0 warn, 0 fail  - operator notification wiring
  evidence       [empty]  0 ok, 0 warn, 0 fail  - local-first evidence ledger and source exporters
  mcp            [empty]  0 ok, 0 warn, 0 fail  - canonical MCP server config sync