The brain registry for AI agents

Publish, discover, verify, benchmark, and deploy complete agent identities. One file. One format. Trusted everywhere.

$ brain pull redclawsystems/kitchen-hostess:latest
✅ Pulled: kitchen-hostess v2.1.0 (4.8KB)
   Trust Score: 92/100 ✅ Verified

$ brain scan kitchen-hostess.brain
🛡️ Tier 1 Static Analysis — 0 critical, 0 high, 0 medium, 1 low
🤖 Agent Integration — Give your agent the skill.md to search, pull, and manage brains programmatically.

What's in a .brain file?

A single, portable archive containing everything an AI agent needs to think, act, and be trusted — from identity and skills to provenance and compliance.

The .brain format is an open standard for packaging complete agent identities. It bundles persona, skills, memory, knowledge, and cryptographic provenance into one distributable file.

Every brain is content-addressed and signed. You can verify who created it, what changed between versions, and whether it passes security analysis — before you ever run it.

Think of it as a container image, but for intelligence instead of infrastructure. One file that works across every runtime, every framework, every deployment target.

agent.brain ├── /meta schema, version, author, license ├── /identity agent.md, soul.md, persona, system prompt ├── /skills tool definitions, MCP configs, workflows ├── /media voice samples, images, avatars ├── /memory vector embeddings, semantic search ├── /knowledge_graph entities, relations ├── /runtime model preferences, resource limits └── /provenance SHA-256 hashes, Ed25519 signatures

Trust is earned, not assumed

Every brain uploaded to the registry is automatically analyzed. Higher tiers unlock deeper verification.

Tier 1 — Static Analysis

Free

67+ patterns. Injection detection. Secrets scanning. Every upload, automatically.

Tier 2 — Behavioral Analysis

Pro

Sandboxed execution. Adversarial probes. Tool monitoring. See what it actually does.

Tier 3 — Compliance

Enterprise

SOC2, HIPAA, PCI-DSS, GDPR mapping. Attestation-ready reports.

✅ Verified 85–100 🔍 Reviewed 65–84 ⚠️ Warning 40–64 🚫 Flagged 0–39

Nine commands. Complete workflow.

From scaffolding to deployment. No config files, no YAML, no ceremony.

$ brain init # Scaffold a new brain project
$ brain pack # Bundle into .brain file
$ brain inspect # Print schema tree & stats
$ brain verify # Check hashes & signatures
$ brain diff a b # Dataset-level comparison
$ brain scan # Tier 1 security analysis
$ brain bench # Run benchmark suites
$ brain push # Upload to registry
$ brain pull org/brain:v1 # Download from registry

Deploy anywhere

From a 3MB binary to a full cloud container. One brain, every target.

🖥️

ZeroClaw

3.4MB binary. Native runtime.

☁️

Cloud VPS

Auto-generated containers.

📱

Edge

OTA push via Tailscale.

🔌

Third-party

LangChain, CrewAI, AutoGen, DSPy.

Sell your brains. Literally.

Publishers earn 85%. Platform handles payments, licensing, and trust verification.

finance-compliance

Regulatory analysis and compliance monitoring for financial institutions.

$49.99/mo ✅ 93/100 ⬇ 3.4k pulls

clinical-copilot

Clinical decision support with HIPAA-compliant data handling.

$99.99/mo ✅ 95/100 ⬇ 6.2k pulls

code-reviewer

Automated code review with security scanning and style enforcement.

$29.99 ✅ 90/100 ⬇ 8.9k pulls

Pricing

Start free. Scale when you're ready.

Free

$0
forever
  • 5 public brains
  • Tier 1 scanning
  • Community support
  • CLI access
Get started

Teams

$9 /user/mo
minimum 5 users
  • Everything in Pro
  • Team management
  • Shared registries
  • Audit logs
  • SSO/SAML
Contact sales

Enterprise

Custom
annual contract
  • Everything in Teams
  • Tier 3 compliance
  • On-prem registry
  • Dedicated support
  • Custom SLAs
Contact sales

Get started

Install the CLI and pull your first brain in under 60 seconds.

$ curl -fsSL https://clawbrainhub.com/install.sh | sh
$ brain pull redclawsystems/kitchen-hostess:latest
$ brain inspect kitchen-hostess.brain
$ brain scan kitchen-hostess.brain
Create your first brain →