Index / Brigade · updated Oct 5, 2026
Brigade
spinabot/brigade · healthy · rank 14 of 50 by stars
Brigade is an enterprise-grade personal AI agent framework that eliminates API key juggling by reusing your existing Claude, ChatGPT, or Copilot subscriptions. Its standout feature is B³ (Brigade Bloody Benchmark), which exposes your agents to the open internet via tunnels to stress-test them against real-world strangers.
Facts
Repository
Runtime
Posture
Community
Security breakdown
Composite 78 / 100 · how these are scored
higher is safer
higher is safer
higher is safer
higher is safer
higher is riskier
Model access
Read from the repository, not written by a model · 67 files examined
5 providers · runs locally · gateway support
Pinned models
gpt-5.4-mini released 2026-03-17 — 7 months old (from the public model catalogue)
Pin last edited 5 months ago
Evidence
Decision
Why choose Brigade over OpenClaw?
Why choose this
- Subscription-based auth removes API key management entirely
- Built-in internet exposure benchmarking (B³) for real-world agent testing
- No telemetry and keys stay local by design
Tradeoffs
- Smaller skill/plugin ecosystem compared to OpenClaw's 5,400+ skills
- Fewer messaging channel integrations documented
- Younger project with less battle-tested stability
Best fit
- Users who want to reuse existing Claude/ChatGPT/Copilot subscriptions without managing API keys
- Developers stress-testing agents against real internet traffic
- Teams needing durable multi-user collaboration mode
Avoid if
- You need a lightweight or minimal-footprint agent runtime
- You want a mature plugin/skill ecosystem like OpenClaw's
- You prefer avoiding tunnel-based public exposure features
Strong evidence from README and active commit history (v1.39.0, frequent releases), but Reddit mentions are entirely unrelated noise (real estate, military brigades), so community sentiment is highly uncertain.
AI layer reviewed Oct 5, 2026 · how this is written
Star activity
11,258 stars today
Overview
Brigade positions itself as an enterprise-grade personal intelligence ecosystem rather than a single app. Its core differentiator is frictionless authentication: instead of juggling API keys, users sign in with existing Claude, ChatGPT, or Copilot subscriptions, or reuse active Claude Code / Codex CLI sessions. API keys remain on the local machine with no telemetry and no middleman proxying.
The headline feature is B³ (Brigade Bloody Benchmark), introduced in v1.9.0, which exposes your agent crew to the open internet through HTTPS tunnels (Cloudflare by default, or self-hosted bore/frp/sish relays) with an invisible secret key gating access. The philosophy is that the only benchmark that matters is real strangers probing your agents at 3 a.m. Recent releases (v1.38–1.39) added a durable team collaboration mode and hardened runtime message pagination, showing active, disciplined engineering with release-please automation and PR-gated main branch.
Compared to OpenClaw, Brigade trades ecosystem breadth for auth simplicity and adversarial testing tooling. OpenClaw offers thousands of community skills and broad messaging-channel coverage, while Brigade focuses on a polished CLI experience, subscription reuse, and real-world exposure benchmarking. Its TypeScript stack means moderate resource footprint, and its tunnel-exposure features, while gated, warrant careful operational review.