brigade-cli 0.28.0__tar.gz → 0.30.0__tar.gz

brigade_cli-0.30.0/PKG-INFO

@@ -0,0 +1,214 @@
+Metadata-Version: 2.4
+Name: brigade-cli
+Version: 0.30.0
+Summary: Your agents run loops. Brigade keeps the receipts. Local control plane for multi-agent coding: share MCP/tools/memory, remember via handoffs, prove with file receipts, improve by promoting skills only on real exit codes. Claude Code, Codex, Grok, Hermes, OpenClaw, and more. Stations for evidence, code graph, pantry, skills. pipx install brigade-cli.
+Author-email: Solomon Neas <me@solomonneas.dev>
+License: MIT
+Project-URL: Homepage, https://brigade.tools
+Project-URL: Documentation, https://brigade.tools/docs
+Project-URL: Compare, https://brigade.tools/compare/sync-mcp-servers-across-coding-agents
+Project-URL: Handoffs, https://brigade.tools/compare/memory-handoffs-across-coding-agents
+Project-URL: CodeGraph, https://brigade.tools/compare/local-code-graph-for-ai-coding-agents
+Project-URL: Evidence, https://brigade.tools/compare/local-evidence-ledger-for-agent-work
+Project-URL: Skills, https://brigade.tools/compare/portable-skills-across-coding-agents
+Project-URL: LLMs, https://brigade.tools/llms.txt
+Project-URL: Repository, https://github.com/escoffier-labs/brigade
+Project-URL: Cookbook, https://escoffierlabs.dev/cookbook/
+Project-URL: CookbookSource, https://github.com/escoffier-labs/solos-cookbook
+Project-URL: OpenClaw, https://github.com/solomonneas/openclaw
+Project-URL: ContentGuard, https://github.com/escoffier-labs/content-guard
+Project-URL: AgentPantry, https://github.com/escoffier-labs/agentpantry
+Project-URL: MiseLedger, https://brigade.tools/miseledger
+Project-URL: GraphTrail, https://brigade.tools/graphtrail
+Project-URL: Issues, https://github.com/escoffier-labs/brigade/issues
+Keywords: agents,ai-agents,agent-memory,agent-handoffs,ai-memory,mcp,mcp-config,mcp-sync,model-context-protocol,openclaw,claude-code,codex,opencode,grok,hermes,memory,bootstrap,brigade,brigade-cli,operator,local-first,guardrails,agents-md,miseledger,graphtrail,agentpantry,code-graph,evidence-ledger
+Classifier: Development Status :: 3 - Alpha
+Classifier: License :: OSI Approved :: MIT License
+Classifier: Programming Language :: Python :: 3
+Classifier: Programming Language :: Python :: 3.10
+Classifier: Programming Language :: Python :: 3.11
+Classifier: Programming Language :: Python :: 3.12
+Classifier: Operating System :: Microsoft :: Windows
+Classifier: Operating System :: MacOS
+Classifier: Operating System :: POSIX :: Linux
+Requires-Python: >=3.10
+Description-Content-Type: text/markdown
+License-File: LICENSE
+Provides-Extra: dev
+Requires-Dist: pytest>=7; extra == "dev"
+Requires-Dist: pytest-cov>=5; extra == "dev"
+Requires-Dist: pytest-xdist>=3.6; extra == "dev"
+Requires-Dist: ruff>=0.15; extra == "dev"
+Requires-Dist: mypy>=1.11; extra == "dev"
+Provides-Extra: grokbot
+Requires-Dist: mcp<3,>=2; extra == "grokbot"
+Requires-Dist: pyjwt[crypto]>=2; extra == "grokbot"
+Provides-Extra: research
+Requires-Dist: playwright>=1.40; extra == "research"
+Dynamic: license-file
+
+<p align="center">
+  <img src="docs/assets/brigade-wordmark.svg" alt="Brigade (by Escoffier Labs)" width="640">
+</p>
+
+<p align="center">
+  One shared record for every coding agent you run: ready work, safe claims, proof that checks passed, and memory that carries over between tools.
+</p>
+
+<p align="center">
+  <a href="https://brigade.tools/docs/getting-started/install">Install</a> ·
+  <a href="https://brigade.tools/docs">Docs</a> ·
+  <a href="docs/comparison.md">Compare</a> ·
+  <a href="QUICKSTART.md">Quickstart</a>
+</p>
+
+<p align="center">
+  <img src="https://shieldcn.dev/github/ci/escoffier-labs/brigade.svg?workflow=ci.yml&branch=main&label=ci&size=xs" alt="CI status">
+  <img src="https://shieldcn.dev/pypi/v/brigade-cli.svg?label=pypi&size=xs" alt="PyPI version">
+  <img src="https://shieldcn.dev/https/raw.githubusercontent.com/escoffier-labs/brigade/download-metrics/downloads.json.svg?size=xs" alt="Combined PyPI and npm downloads per month">
+  <img src="https://shieldcn.dev/badge/python-3.10+-blue.svg?logo=python&logoColor=white&size=xs" alt="Python 3.10+">
+  <img src="https://shieldcn.dev/badge/rust-code_graph-b7410e.svg?logo=rust&logoColor=white&size=xs" alt="Rust: code graph engine">
+  <img src="https://shieldcn.dev/badge/go-evidence_log-00add8.svg?logo=go&logoColor=white&size=xs" alt="Go: evidence log engine">
+  <img src="https://shieldcn.dev/badge/license-MIT-4e7247.svg?size=xs" alt="MIT license">
+</p>
+
+Brigade is a local CLI for people who point Claude Code, Codex, Cursor, and other agents at the same repos. Each harness keeps its own notes, its own MCP config, and its own idea of what "tests pass" means. Brigade gives them one ledger in the repo instead: what is ready, who claimed it, which commands actually ran and how they exited, and what the last session handed off.
+
+**Stable:** v0.29.0 (`brigade-cli==0.29.0` on PyPI, `@escoffier-labs/brigade@0.29.0` on npm). **Beta:** main is the 0.30 preview line. See the [changelog](CHANGELOG.md) for what each release changed.
+
+The downloads badge sums `brigade-cli` on PyPI (excluding mirrors) and `@escoffier-labs/brigade` on npm over the same 30 completed UTC days. Downloads include automation and do not count unique people. The [metric JSON](https://raw.githubusercontent.com/escoffier-labs/brigade/download-metrics/downloads.json) records each source, the period, and the last successful refresh. npm platform companions are excluded. Daily refreshes retain the previous metric if a source fails validation or its latest day is more than three UTC days old.
+
+## What it does
+
+- **Ready work and safe claims.** `brigade work ready` lists unblocked tasks. Claims are atomic, so two agents cannot take the same item.
+- **Verification receipts.** A check run through Brigade records the command, the real exit code, and the Git state. An agent's "tests pass" becomes something you can open and read.
+- **Shared memory with an owner.** Agents write handoffs. The memory owner lints and files them. Safe notes can auto-file, and risky ones wait for review. Session-start recall is capped.
+- **One tool catalog, previewed first.** MCP servers, tools, and skills are defined once and shown as a dry run before Brigade writes them into each harness.
+- **Cloud agents tracked next to local ones.** Tasks you hand to Codex Cloud, Cursor cloud agents, or Jules are registered with their provider task ID and reconciled against provider and GitHub state, so they show up beside local runs instead of disappearing into another tab.
+- **Optional fleet view.** A self-hosted hub can arbitrate repo claims across machines and show runs, harness sessions, and cloud tasks on one board.
+
+Learning in Brigade means scoring skills against captured receipts, then promoting or rolling them back. It is not autonomous reflective memory.
+
+<details>
+<summary><b>Examples</b></summary>
+
+**Two agents, one backlog.** Each agent asks for the next ready task. The second claim on the same item fails with exit 13 and names the holder.
+
+```bash
+brigade work ready --parallel-safe          # ready tasks, grouped into waves that do not touch the same files
+brigade work claim --next --actor codex     # take the highest-priority ready task
+```
+
+**"Tests pass" with a receipt.** The agent runs the check through Brigade instead of reporting it from memory.
+
+```bash
+brigade work verify run --target . --command "pytest -q"
+```
+
+The receipt keeps the command, exit code, Git commit, and dirty state, so a reviewer can check the claim without rerunning it.
+
+**Switching harnesses mid-task.** Claude Code ends a session with a handoff. The next Codex session starts from the brief instead of a blank context.
+
+```bash
+brigade handoff lint                        # check the handoff before it is filed
+brigade work brief --target .               # what the next agent sees at session start
+```
+
+**One MCP config for every tool.**
+
+```bash
+brigade mcp plan                            # read-only preview
+brigade mcp sync --write                    # merge into each harness's native config
+```
+
+</details>
+
+## Install
+
+If Node.js 22 or newer and npm are installed on glibc Linux x64/ARM64, macOS x64/ARM64, or Windows x64, install the [npm package](https://www.npmjs.com/package/@escoffier-labs/brigade). This route does not need Python or pipx first:
+
+```bash
+npm install --global @escoffier-labs/brigade@0.29.0
+brigade setup
+```
+
+To try it without a global install, run each command through npx instead, for example `npx --yes @escoffier-labs/brigade@0.29.0 --version`. The package brings its own uv, through platform packages that npm installs as optional dependencies, and the published 0.29.0 wheel. Leave optional dependencies enabled. The first launch downloads a managed Python, so it needs network access. Install Git yourself for repository features. npm targets are Linux (glibc) x64 and ARM64, macOS x64 and ARM64, and Windows x64. Upgrade with npm, because `brigade update` refuses npm-managed installs. Platform details and known limits: [npm package notes](npm/feasibility/README.md).
+
+On other platforms, or to use Python 3.10 or newer directly:
+
+```bash
+pipx install brigade-cli      # or: uv tool install brigade-cli
+brigade setup
+```
+
+Then follow the [install guide](https://brigade.tools/docs/getting-started/install) or [QUICKSTART.md](QUICKSTART.md). To let your coding agent do the setup, hand it [first 10 minutes](docs/first-10-minutes.md) and the [agents guide](docs/agents-guide.md).
+
+For preview builds on a pipx install, run `brigade update --channel beta`. Details: [update channels](docs/update-channels.md).
+
+## How it compares
+
+Most neighbors go deeper on one layer. Brigade connects work, memory, tool config, and evidence around the same coding-agent task. Where a neighbor is stronger, its cell says so. Sources, dated notes, and projects that left these tables: [docs/comparison.md](docs/comparison.md) (checked 2026-10-18).
+
+Key: <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Supported · <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Partial · <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented
+
+### Work and orchestration
+
+| Project | Ready work | Claim safety | Parallel and run control | Task store | Verification |
+| --- | --- | --- | --- | --- | --- |
+| [Brigade](https://github.com/escoffier-labs/brigade) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Ready set + parallel-safe waves | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Fail-closed claims, optional hub | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Waves, campaigns, declared limits | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Local JSON ledger, not a task DB | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Command receipts |
+| [Beads](https://github.com/gastownhall/beads) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Built in | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Atomic claim | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Agent workflow + Dolt sync | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Dolt-backed distributed store | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Task history |
+| [Gas Town](https://github.com/gastownhall/gastown) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Beads-backed | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Multi-agent runtime | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Beads ledger | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Refinery merge-queue gates |
+| [Claude Code agent teams](https://code.claude.com/docs/en/agent-teams) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Shared list with dependencies | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> File-locked self-claim | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Lead + teammates, experimental | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> One Claude Code session | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> TaskCompleted hook can block |
+| [Backlog.md](https://github.com/MrLesk/Backlog.md) | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Board + dependencies | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Markdown in Git | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Acceptance criteria |
+
+### Memory
+
+| Project | Durable memory | Capture and recall | Care and operations | Learning | Provenance |
+| --- | --- | --- | --- | --- | --- |
+| [Brigade](https://github.com/escoffier-labs/brigade) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Shared files + cards | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Linted handoffs + bounded recall | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Owner policy + Memory Operations | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Outcome score, promote, rollback | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Work receipts + sources |
+| [agentmemory](https://github.com/rohitg00/agentmemory) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Shared coding-agent memory | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Hooks + MCP + REST | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Decay, audit, governed delete | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Lessons + reflect | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Write-time origin + citations |
+| [Hindsight](https://github.com/vectorize-io/hindsight) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Facts + experiences + models | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Retain + recall | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Reflect + feedback | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Reflective memory learning | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Temporal + causal links |
+| [Mem0](https://github.com/mem0ai/mem0) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> User + agent + session | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Extraction + hybrid search | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> History + metadata | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Extraction + retrieval | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Timestamps + entities |
+| [Built-in harness memory](https://code.claude.com/docs/en/memory) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Per tool | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Rules + auto memory | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Varies by harness | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Harness-specific | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Harness-specific |
+
+### Tools and configuration
+
+| Project | MCP | Rules and skills | Projection | Source of truth | Audit and rollback |
+| --- | --- | --- | --- | --- | --- |
+| [Brigade](https://github.com/escoffier-labs/brigade) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> MCP + tools catalog | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Skills + harness adapters | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Dry run first, then write | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Brigade catalog | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Preserve, gate, roll back |
+| [Rulesync](https://github.com/dyoshikawa/rulesync) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Supported | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Rules, skills, hooks, subagents | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Per-target generation | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Rulesync project | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Generated outputs |
+| [Ruler](https://github.com/intellectronica/ruler) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Supported | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Rules, skills, subagents | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Per-agent files | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> `.ruler/` directory | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Gitignore automation, no rollback |
+| [add-mcp](https://github.com/neon-solutions/add-mcp) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> MCP servers | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Maps fields into native files | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Native files + sync | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Warns on dropped fields |
+| [agentsync](https://agentsync.cc/) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Many agent targets | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Skills, hooks, memory, plugins | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Native/lossy/skipped report | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Canonical config | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Native/lossy/skipped report |
+
+### Evidence and operations
+
+| Project | Receipts | Code impact | Agent activity | Cloud state | Replay and learning |
+| --- | --- | --- | --- | --- | --- |
+| [Brigade](https://github.com/escoffier-labs/brigade) | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Command + exit code + Git state | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Impact queries + blast radius | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Center, fleet board, harness sessions | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Codex, Cursor, Jules task status | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Score + rollback, run diff/resume |
+| [ActiveGraph](https://github.com/yoheinakajima/activegraph) | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Event trace | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Reactive behaviors | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Replay, fork, and diff |
+| [Graphiti](https://github.com/getzep/graphiti) | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Episode provenance | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Temporal fact updates |
+| [Codex Cloud](https://developers.openai.com/codex/cloud) | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Changed files + test results | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented | <img src="docs/assets/status-partial.svg" width="10" height="10" alt="Partial"> Task review | <img src="docs/assets/status-supported.svg" width="10" height="10" alt="Supported"> Hosted task workspaces | <img src="docs/assets/status-undocumented.svg" width="10" height="10" alt="Not documented"> Not documented |
+
+"Not documented" means the capability was not found in that project's official README or docs on the dated pass. It is not proof of absence.
+
+## Scope
+
+Brigade is a local, file-first control plane. Nothing runs until you run a command or your own scheduler does. The work ledger is machine-local JSON, not a distributed task database, and the optional fleet hub coordinates claims and events without replacing local records. Run budgets apply only when declared. Brigade does not push to GitHub, publish packages, or send messages unless you use an explicit send action. Full boundaries: [execution model](docs/execution-model.md).
+
+Name collisions: this is `brigade-cli` from [escoffier-labs/brigade](https://github.com/escoffier-labs/brigade), not the archived CNCF Kubernetes Brigade, Spinabot Brigade, or the 2017 Python package that became Nornir.
+
+## Docs
+
+- [First 10 minutes](docs/first-10-minutes.md) · [Overview](docs/overview.md) · [Technical guide](docs/technical-guide.md) · [Comparison](docs/comparison.md)
+- [Execution model](docs/execution-model.md) · [Work closeout](docs/work-closeout.md) · [Outcome scoring](docs/outcome-scoring.md)
+- [Memory care](docs/memory-care.md) · [Memory operations](docs/memory-operations.md) · [Code intelligence](docs/code-intelligence.md) · [MCP sync](docs/mcp-sync.md)
+- [Agents guide](docs/agents-guide.md) · [Agent-assisted setup](docs/agent-assisted-setup.md) · [README coverage](docs/readme-coverage.md)
+- [Runbooks](docs/runbooks/README.md) · [Fleet hub on a Proxmox LXC](docs/runbooks/fleet-hub-proxmox.md)
+- [Security](docs/security.md) · [Command inventory](docs/command-inventory.md) · [Contributing](CONTRIBUTING.md) · [Roadmap](ROADMAP.md)
+
+## License
+
+MIT. See [LICENSE](LICENSE).
+
+Project identity: GitHub [`escoffier-labs/brigade`](https://github.com/escoffier-labs/brigade), website [brigade.tools](https://brigade.tools), PyPI [`brigade-cli`](https://pypi.org/project/brigade-cli/), npm [`@escoffier-labs/brigade`](https://www.npmjs.com/package/@escoffier-labs/brigade), command `brigade`.

brigade_cli-0.30.0/QUICKSTART.md

@@ -0,0 +1,367 @@
+# Quickstart
+
+Five minutes from install to a working agent kitchen.
+
+## 1. Install
+
+Brigade supports Linux, macOS, and Windows. Pick one route: npm on glibc Linux x64/ARM64, macOS x64/ARM64, or Windows x64 if Node.js 22+ and npm are already installed, or Python 3.10 or newer with pipx or uv. Other targets, including musl Linux (such as Alpine) and Windows ARM64, use the Python route.
+
+### npm
+
+The npm package needs Node.js 22 or newer and npm. It does not need Python or pipx first:
+
+```bash
+npm install --global @escoffier-labs/brigade@0.29.0
+brigade --version
+brigade setup
+```
+
+For a one-off run without a global install, put npx in front of every Brigade command. No global `brigade` command exists on this route:
+
+```bash
+npx --yes @escoffier-labs/brigade@0.29.0 --version
+npx --yes @escoffier-labs/brigade@0.29.0 setup
+```
+
+The first launch downloads a managed Python, so it needs network access. Git is not included. Install it yourself for repository features. Supported npm targets are Linux (glibc) x64 and ARM64, macOS x64 and ARM64, and Windows x64. Other platforms use the Python route below. Keep npm optional dependencies enabled, because the uv binary arrives as a platform package. If pipx or uv also installed Brigade, check which `brigade` comes first on PATH.
+
+Upgrade by installing a newer exact version with npm. `brigade update` refuses npm-managed installs. Service units installed from an npm install point at that version's private Python, so reinstall them after each npm upgrade. The [npm package notes](npm/feasibility/README.md) cover the cache location, offline behavior, and known platform limits.
+
+### Python with pipx
+
+Install `pipx` with the package manager for your OS, then install Brigade.
+
+#### Linux
+
+Ubuntu 23.04+, Debian 12+, and Fedora use an externally managed system Python. Install `pipx` from the distribution package instead of using `pip install --user`:
+
+```bash
+# Ubuntu or Debian
+sudo apt update
+sudo apt install pipx
+
+# Fedora
+sudo dnf install pipx
+
+pipx ensurepath
+```
+
+##### Arch Linux and Omarchy
+
+On Arch Linux, install [the `python-pipx` package](https://archlinux.org/packages/extra/any/python-pipx/):
+
+```bash
+sudo pacman -Syu --needed python-pipx
+pipx ensurepath
+```
+
+This Arch command performs a full system upgrade and installs `pipx`.
+
+On Omarchy, use its [package command](https://omarchy.org/manual/other-packages/):
+
+```bash
+omarchy pkg add python-pipx
+pipx ensurepath
+```
+
+Use `omarchy update` for [Omarchy system updates](https://omarchy.org/manual/updates/) so its snapshots, migrations, and configuration updates run with the package upgrade.
+
+On either distribution, open a new terminal, then follow the Brigade install commands below. Run `pipx` as your regular user. It keeps Brigade isolated from the externally managed system Python.
+
+After an Arch upgrade replaces Python with a new minor version, an existing pipx environment may need rebuilding:
+
+```bash
+pipx reinstall brigade-cli
+brigade --version
+```
+
+On another distribution, use its `pipx` package when available. The [pipx install guide](https://pipx.pypa.io/latest/how-to/install-pipx.html) covers the virtual-environment fallback.
+
+#### macOS
+
+```bash
+brew install pipx
+pipx ensurepath
+```
+
+#### Windows PowerShell
+
+```powershell
+scoop install pipx
+pipx ensurepath
+```
+
+Without Scoop, use Python's Windows launcher:
+
+```powershell
+py -m pip install --user pipx
+# Run pipx.exe ensurepath from the directory printed by pip if it is not on PATH yet.
+```
+
+Open a new terminal after `pipx ensurepath`, then continue on any OS:
+
+```bash
+pipx install brigade-cli
+brigade setup
+```
+
+If you already use uv, install the same package with:
+
+```bash
+uv tool install brigade-cli
+brigade setup
+```
+
+For an intentional Brigade development machine with a pipx install, follow the preview line:
+
+```bash
+brigade update --channel beta
+```
+
+Rollback to stable:
+
+```bash
+brigade update --channel stable --switch-channel
+```
+
+## First install
+
+The canonical first-run command is `brigade operator quickstart`. It installs the template files, wires the operator config, scaffolds the MCP catalog and dogfood/work-loop config, and runs the health checks in one shot:
+
+```bash
+# Code repo with Codex as the writer
+brigade operator quickstart --target ./my-repo --harnesses codex
+
+# OpenClaw or Hermes workspace instead of a code repo
+brigade operator quickstart --target ~/agent-workspace --depth workspace --harnesses openclaw,hermes --owner openclaw
+```
+
+Pass `--dry-run` first to preview the planned steps without writing anything.
+
+PowerShell uses Windows path spelling when you provide a path explicitly:
+
+```powershell
+brigade operator quickstart --target .\my-repo --harnesses codex
+brigade operator doctor --target .\my-repo --profile local-operator
+```
+
+No WSL is required. Worker CLIs must resolve to native executables; when an npm tool resolves only to a `.cmd` or `.bat` shim, Brigade reports the shim and asks for a native launcher. Shell completions currently cover bash, zsh, and fish. `brigade stations verify` is POSIX-only and reports `unsupported-platform` on Windows; quickstart, doctor, the work loop, and native worker runs are supported.
+
+Two commands share this surface: `brigade init` installs the template files only, and `brigade operator quickstart` wraps it with operator config, the MCP and dogfood on-ramps, writer verification, and health checks. Use `init` when you want the interactive harness picker or just the files:
+
+```bash
+$ brigade init --target ~/agent-kitchen
+
+Which harnesses do you use? (type numbers separated by space/comma to toggle, enter to confirm)
+  [x] 1. Claude Code
+  [ ] 2. Codex
+  [ ] 3. OpenCode
+  [ ] 4. Antigravity
+  [ ] 5. Pi
+  [ ] 6. Cursor
+  [ ] 7. Aider
+  [ ] 8. Goose
+  [ ] 9. Continue
+  [ ] 10. GitHub Copilot CLI
+  [ ] 11. Qwen Code
+  [ ] 12. Kimi Code
+  [ ] 13. AdaL
+  [ ] 14. OpenHands
+  [ ] 15. Grok CLI
+  [ ] 16. Amp
+  [ ] 17. Crush
+  [ ] 18. OpenClaw
+  [ ] 19. Hermes
+
+Depth? (type a number, enter for default)
+  * 1. repo       (handoff flow + publish guard)
+    2. workspace  (full home: MEMORY.md, TOOLS.md, USER.md, ...)
+
+Add-ons? (type numbers separated by space/comma to toggle, enter to confirm)
+  [ ] 1. publisher  (content-guard policies for blog/social/docs)
+```
+
+Defaults are claude harness, repo depth, no includes. Enter ships the install.
+
+## CI / scripted install
+
+Pass flags directly to skip the prompt. The same flags work on `operator quickstart`:
+
+```bash
+# Claude Code + Codex + OpenClaw, full workspace
+brigade operator quickstart --target ~/agent-kitchen \
+  --depth workspace \
+  --harnesses claude,codex,openclaw
+
+# Codex-only project, minimal install
+brigade operator quickstart --target ./my-project --depth repo --harnesses codex
+
+# Template files only, no harness-specific files
+brigade init --target ./my-project --harnesses none
+```
+
+## Verifying
+
+After install, check the operator profile:
+
+```bash
+brigade operator doctor --target <path> --profile local-operator
+```
+
+For the file-by-file view, `brigade doctor --target <path>` reports the apparent harness shape and checks every configured inbox and adapter:
+
+```
+brigade doctor: target /home/you/agent-kitchen
+  harnesses: claude, codex, openclaw (owner=openclaw, depth=workspace)
+  [ok]   bootstrap: AGENTS.md   /home/you/agent-kitchen/AGENTS.md
+  [ok]   handoff: claude inbox  /home/you/agent-kitchen/.claude/memory-handoffs
+  [ok]   handoff: codex inbox   /home/you/agent-kitchen/.codex/memory-handoffs
+  [ok]   openclaw: config        /home/you/.openclaw/openclaw.json
+  ...
+```
+
+A `[fail]` line means the install is incomplete; `[warn]` is informational; `[todo]` means the check needs your attention.
+
+## Reconfiguring
+
+To change which harnesses are installed on an existing target:
+
+```bash
+# Add a harness
+brigade reconfigure --target . --harnesses claude,codex
+
+# Drop one (without removing its files)
+brigade reconfigure --target . --harnesses claude
+
+# Drop one and remove its files
+brigade reconfigure --target . --harnesses claude --prune
+```
+
+## The handoff flow
+
+The starter handoff template lives at `<inbox>/TEMPLATE.md`. Copy it to a new dated file (e.g. `2026-05-16-1430-fixed-X.md`), fill it in, and the ingester promotes safe card handoffs into `memory/cards/`, appends targeted updates to the right file, and kicks ambiguous material to the review inbox.
+
+When Claude Code is a selected writer harness, commit the managed `.claude/memory-handoffs/TEMPLATE.md` so the handoff format travels with the repository and Claude onboarding can verify it. Other files under `.claude/` and session handoff notes in that inbox stay local-only. The same tracked-template pattern applies to other writer inboxes. If a global Git exclude hides the Claude template, use the repo-local recovery steps in [docs/new-user-quickstart.md](docs/new-user-quickstart.md#claude-handoff-template-hidden-by-a-global-git-exclude). Brigade never edits global Git configuration.
+
+See the [Solo Cookbook](https://github.com/escoffier-labs/solos-cookbook) for the longer-form guidance on what makes a good handoff and when to use which routing.
+
+## Your daily loop
+
+Handoffs are one half. The other half is routing your actual work through Brigade so it produces a real signal, instead of leaving Brigade installed-but-dormant. `brigade init` wires a `brigade-work` skill into each harness so your agent does this automatically; by hand it is three steps:
+
+```bash
+brigade work brief --target .                                  # 1. what's pending (and whether the loop is being fed)
+brigade work verify run --target . --command "pytest -q" --capture <skill-or-card>   # 2. verify + capture in one step
+# 3. write a Memory Handoff for anything durable; the ratchet (outcome reconcile) runs hands-off on your cron
+```
+
+Capture against an id you actually have: a skill you followed, a memory card (`--kind card`), or `brigade-work` itself when nothing else applies. `brigade work brief` surfaces the loop's own health so an empty ledger never goes unnoticed.
+
+The `--command` example captures audit evidence only. Artifact scoring requires an eligible receipt from `--manifest <id>` using a tracked verifier under `verify/manifests/`. Inspect the reported scoreability before relying on it for promotion or rollback.
+
+## Optional: close the GraphTrail ↔ Brigade ↔ MiseLedger loop
+
+0.21.0 already ships receipt deltas, MiseLedger export, evidence briefs, and context evals. Productizing them is a short dogfood path:
+
+```bash
+# optional stations (fail-open everywhere if absent)
+brigade setup                    # run once: verified GraphTrail, graphtrail-mcp, MiseLedger, and SessionFind
+brigade code sync .             # explicitly runs local GraphTrail and builds .graphtrail/graphtrail.db
+brigade code context "auth receipt flow"
+brigade code impact brigade.work.verify.run
+
+# evidence station CLI: explicit local operations
+brigade evidence crawl sessions
+brigade evidence search "auth receipt flow"
+brigade evidence crawl plan     # review-only miseledger init + crawl commands
+brigade evidence doctor
+brigade evidence export plan    # review-only receipts export path
+
+# one operator glance: doctors + graph ok / ledger ok / last brief hit rate
+brigade operator checkup --target .
+
+# after real work flows through verify/run + capture:
+brigade receipts export miseledger --target . --new-only --import
+brigade outcome rank --target .    # surfaces brief_hit as a skill quality signal
+# next brigade run attaches a capped MiseLedger evidence brief automatically
+```
+
+One-release compatibility note: `brigade add evidence`, `brigade add search`, and `brigade add graphtrail` remain available for independent installations. A Brigade-managed install needs only `brigade setup` before these facade commands.
+
+That is the differentiated loop: receipts that feed the next run's context, with a measured hit rate (`context_eval.brief_hit_rate`) on whether the pre-run brief named the files the run actually touched.
+
+## Optional: multi-machine pantry (Agent Pantry)
+
+When the agent runs on a different machine from your daily browser login, the **pantry** station wires [Agent Pantry](https://brigade.tools/agentpantry) as a process-boundary Go binary. Brigade installs and plans; it does not mint PSKs or start services.
+
+```bash
+brigade extras on                                         # enables pantry and notifications below
+brigade add pantry                                          # go install agentpantry
+brigade pantry setup plan --role sink --peer 127.0.0.1:8787 # agent host
+brigade pantry setup plan --role source --peer <sink>:8787  # daily driver
+# run the printed agentpantry init/keygen/source|sink commands yourself
+brigade pantry doctor
+brigade pantry expiry-alert                                 # preview near-expiry cookies
+# optional notify path:
+brigade add notifications
+brigade pantry expiry-alert --send
+```
+
+`brigade pantry status` and `doctor` always print a `next:` block so the path stays obvious after install.
+
+## Optional: search and tokens stations
+
+```bash
+# GraphTrail is installed by `brigade setup`; use its facade commands.
+brigade code sync .             # preferred GraphTrail facade
+brigade code context "auth receipt flow"
+brigade code impact brigade.work.verify.run
+# `brigade search sync|context|impact` remain compatibility aliases.
+brigade search sync plan        # review-only; does not run GraphTrail
+brigade search doctor
+
+# Token Glace (output compaction; TokenJuice was the old name) + optional usage export
+brigade add tokens
+brigade tokens wire plan
+brigade tokens doctor
+```
+
+## Optional: discover external station catalogs
+
+Sidecars can ship a `station.json` (`schema: brigade.station.v1`). Discover them locally without folding them into the core:
+
+```bash
+brigade stations list
+brigade stations discover --root ~/repos
+brigade stations verify /path/to/sidecar          # read-only conformance from an explicit local path
+brigade stations verify /path/to/sidecar --json   # bounded result; no raw child output
+brigade add /path/to/sidecar --install   # runs printed install args after review
+```
+
+`stations verify` never installs the sidecar. On POSIX it runs declared read-only commands or narrowly checked help/version probes with an isolated temporary home, finite positive timeouts, process-group cleanup, and a 64 KiB combined stdout/stderr ceiling. Windows returns `unsupported-platform` before spawning a probe. Use `--check-managed` in coordinated fleet checks to turn matching Brigade catalog drift from an advisory into a failure. Legacy v1 manifests remain discoverable, but missing bounds fail strict verification. Embedded, deprecated, and historical manifests report a lifecycle skip without running external commands.
+
+## Optional: operator notifications (extras)
+
+`notifications` and `pantry` are part of the **extras** surface. Enable once with `brigade extras on` (or `BRIGADE_EXTRAS=1` per invocation). Core station CLIs (`evidence`, `search`, `tokens`) always register.
+
+```bash
+brigade extras on
+brigade add notifications
+brigade notifications status
+brigade notifications setup plan --profile operator
+# Brigade never sends from doctor/status; secrets stay in env vars referenced by config.toml
+```
+
+## Extras honesty
+
+`brigade --help` shows the core pitch by default. The wider operator suite (release trains, fleet health, research, pantry, notifications, ...) stays fully functional but only registers when extras are on. That keeps command count honest for new users without deleting capability for operators.
+
+## Next steps
+
+- Read [the cookbook](https://github.com/escoffier-labs/solos-cookbook) for the deep version of every concept here.
+- Customize `USER.md` and `TOOLS.md` with your real preferences and runbooks (kept private; do not commit personal details).
+- Wire the ingester on a cron or a manual end-of-day workflow.
+- Add a memory-care staleness scan when your card set starts to matter. See [docs/memory-care.md](docs/memory-care.md).
+- If you use Token Glace (formerly TokenJuice), wire Claude Code and Codex hooks deliberately and tell agents what the wrapper means. See the tokens station in [docs/technical-guide.md](docs/technical-guide.md#managed-stations).
+- Run `brigade work bootstrap` inside active repos that did not use quickstart when you want the dogfood-backed daily work loop, scanner inbox, and local evidence receipts.
+- Wire Grok as a writer harness with `--harnesses ...,grok` so handoffs land in `.grok/memory-handoffs/` (first-class, same as Claude/Codex).