Search BrigadeStack

Search BrigadeStack pages and docs.

BrigadeStack
Menu
Docs menu

Docs / Reference

tools

The portable tool catalog, call review, runtimes, and execution policy.

Use this to keep one reviewed catalog of the tools your agents can call, with approval before anything runs.

Catalog discovery

  • brigadestack tools init writes gitignored .brigade/tools.toml with local examples for portable slash commands and superpowers.
  • brigadestack tools list, show <tool-id>, and search <query> inspect logical tool entries across source families such as skill, slash-command, superpower, mcp, openapi, graphql, script, and custom.
  • brigadestack tools describe <tool-id> and brigadestack tools contracts inspect schema-backed call contracts, permissions, effects, approval mode, env labels, and argument templates.
brigadestack tools init
brigadestack tools list
brigadestack tools show simplify
brigadestack tools search simplify
brigadestack tools describe simplify
brigadestack tools contracts

Call planning and approval

  • brigadestack tools call plan <tool-id> --args ... validates local JSON args against the configured input schema and returns a redacted wrapper-friendly call plan without executing the tool.
  • brigadestack tools call queue/list/show/approve/reject/hold stores planned calls in .brigade/tools/calls.jsonl for local review. Approval changes status only and never executes a tool.
  • brigadestack tools call run <call-id> and brigadestack tools call run --next execute approved, unblocked local script calls and approved local mcp calls, then write local receipts and stdout/stderr logs under .brigade/tools/runs/.
brigadestack tools call plan simplify --args '{"path":"README.md"}'
brigadestack tools call queue simplify --args '{"path":"README.md"}'
brigadestack tools call list
brigadestack tools call show <call-id>
brigadestack tools call approve <call-id>
brigadestack tools call reject <call-id> --reason "not needed"
brigadestack tools call hold <call-id> --reason "needs review"
brigadestack tools call run <call-id>
brigadestack tools call run --next

Run history and checkpoints

  • brigadestack tools run list/show/latest inspects local execution receipts, and brigadestack tools run replay <run-id> creates a new pending reviewed replay candidate after revalidating current contract, source, runtime, and policy state. Replay never reruns directly.
  • brigadestack tools checkpoint list/show/approve/reject/resume reviews script-requested local checkpoints under .brigade/tools/checkpoints/. Resume requires explicit checkpoint approval and revalidates runtime and policy gates.
brigadestack tools run list
brigadestack tools run show <run-id>
brigadestack tools run latest
brigadestack tools run replay <run-id>
brigadestack tools checkpoint list
brigadestack tools checkpoint show <checkpoint-id>
brigadestack tools checkpoint approve <checkpoint-id> --choice continue
brigadestack tools checkpoint reject <checkpoint-id> --reason "not safe"
brigadestack tools checkpoint resume <checkpoint-id>

Runtimes and policy

  • brigadestack tools runtime init/list/show/status/start/stop/restart/doctor manages explicit local runtimes used by portable tool calls, writing PID files and logs under .brigade/tools/runtime/.
  • brigadestack tools policy init/show/doctor manages host-local execution policy, including allowed effects, timeout caps, runtime allow-lists, approval modes, and environment label bindings.
brigadestack tools runtime init
brigadestack tools runtime list
brigadestack tools runtime status
brigadestack tools runtime start <runtime-id>
brigadestack tools runtime stop <runtime-id>
brigadestack tools runtime doctor
brigadestack tools policy init
brigadestack tools policy show
brigadestack tools policy doctor

Projections, apply, and doctor

  • brigadestack tools plan previews exact projection creates, updates, skips, unmanaged conflicts, and local-edit conflicts for all configured agent tool targets.
  • brigadestack tools apply <tool-id> and brigadestack tools apply --all explicitly write managed agent tool projections. Use --dry-run to preview writes and --force only to overwrite unmanaged or locally edited projection files.
  • brigadestack tools doctor reports missing sources, manifests, schemas, invalid contracts, missing examples, bad argument templates, projections, unmanaged projections, locally edited managed projections, stale projection fingerprints, MCP config issues, stale health files, unsafe auth/env field names, and high-risk command shapes.
  • brigadestack tools import-issues turns catalog health issues into local tool-catalog work imports with stable fingerprints and dismiss-until-changed behavior.

Built-in tool projections from quickstart and brigadestack tools defaults are scoped to the workspace’s selected agent tools plus the neutral scripts/ folder, so a --harnesses codex setup does not get folders for every supported agent tool. The managed .gitignore block written by tools defaults, tools init, and tools pack import covers the handoff-inbox entries for every selected agent tool.

brigadestack tools plan
brigadestack tools plan simplify
brigadestack tools apply simplify --dry-run
brigadestack tools apply simplify
brigadestack tools apply --all
brigadestack tools doctor
brigadestack tools import-issues

Projection parity and sync

  • brigadestack tools parity status reports projection-parity closeout state, and brigadestack tools parity closeout reviews current projection parity issues. Use --reason to record a review or defer reason and --defer to mark issues deferred instead of reviewed.
  • brigadestack tools sync plan plans reviewed add-only projection sync without writing, and brigadestack tools sync apply applies it. Use --write to write reviewed add-only projections, --all to apply all configured tool projections, --dry-run to plan writes without changing files, and --force to allow intentional overwrites through managed apply.
brigadestack tools parity status
brigadestack tools parity closeout --reason "reviewed"
brigadestack tools parity closeout --defer --reason "later"
brigadestack tools sync plan
brigadestack tools sync apply --write
brigadestack tools sync apply --all --write
brigadestack tools sync apply --dry-run

brigadestack tools apply writes projections. brigadestack tools call run executes approved local script entries and approved local mcp entries with already-running managed runtimes. BrigadeStack does not connect to remote MCP servers, fetch OpenAPI or GraphQL schemas, store auth, or send approval notifications. Keep tokens, secrets, private URLs, and host-private paths out of public catalog templates.