About Behavior Runtime MCP
Behavior Runtime MCP is an in-browser behavior-tree execution engine for AI agents, with deterministic operation at about 60 Hz, reactive triggers, and audit records. It provides tools for runtime state, blackboards, and behavior management, alongside layered safety controls.
Key features
- 01In-browser behavior-tree execution at about 60 Hz
- 02Reactive triggers with priority preemption
- 03Capture and inspect deterministic replays
- 04Query runtime status, metrics, and blackboard
- 05Manage behavior trees and Merkle audits
- 06Safety limits and emergency halt
Use cases
- 01Run AI-agent behavior trees in a browser runtime
- 02Interrupt or pause behavior on triggers
- 03Inspect runtime state and blackboard variables
- 04Audit execution records and restore checkpoints
MCP support
The MCP server exposes tools to load and control behavior trees, register triggers, inspect runtime status and metrics, and manage blackboards and the runtime database.
Configuration
Local server config is the same for every client. Copy it into your client’s MCP config file.
Tools
10Saved tool list · updated 2w ago
| Parameter | Type | Required | Description |
|---|---|---|---|
| action | enum | Yes | Behavior loading operation: load, unload, swapload | unload | swap |
| project | string | No | Target project slug |
| trace_id | string | No | Distributed trace ID for cross-server correlation |
| intention | object | No | Optional linked intention payload matching the dispatch token |
| parameters | object | No | Initial execution parameters |
| session_id | string | No | Linked state-memory session ID |
| intention_id | string | No | Linked agent-reasoning intention ID |
| behavior_name | string | Yes | Name of the behavior tree to load |
| dispatch_token | object | No | Cryptographically signed dispatch token issued by agent-reasoning-mcp (gate_intention) |
| dispatch_token.aud | string | No | |
| dispatch_token.token_id | string | No | |
| dispatch_token.issued_at | string | No | |
| dispatch_token.expires_at | string | No | |
| dispatch_token.params_hash | string | No | |
| dispatch_token.intention_id | string | No | |
| dispatch_token.behavior_name | string | No | |
| dispatch_token.hmac_signature | string | No | |
| behavior_version | number | No | Version of behavior tree (defaults to latest) |
| client_request_id | string | No | Idempotency key |
Details
From the project README@putervision/behavior-mcp
High-Frequency (~60Hz) In-Browser Behavior Tree Tactical Runtime Engine for AI Agents
@putervision/behavior-mcp is a Model Context Protocol (MCP) server that executes deterministic behavior trees directly in browser runtimes at ~60Hz with reactive trigger preemption, frame recordings, SHA-256 Merkle audit chains, and a 5-layer safety guardrail stack.
🌐 Official Documentation: putervision.com • Interactive Web Docs
⚡ 15-Second Quick Start
# 1. Initialize runtime & seed default behavior trees in your workspace
npx @putervision/behavior-mcp init
# 2. Run system diagnostic and health checks
npx @putervision/behavior-mcp doctor
# 3. Inspect active executions, blackboard state, and reactive triggers
npx @putervision/behavior-mcp inspect
🛠️ 10 Core MCP Tools
| Tool | Actions | Purpose |
|---|---|---|
load_behavior | load, unload, swap | Inject, initialize, or hot-swap a behavior tree in browser runtime |
set_parameters | set, get, reset | Dynamically update or query runtime execution parameters |
get_status | current, history, tree_state | Query active status, node traversal path, tick count, duration |
abort_behavior | abort, pause, resume | Halt, pause, or resume behavior execution safely |
register_trigger | register, list | Configure reactive interrupt triggers with priority preemption |
replay_recording | capture, list | Capture or inspect deterministic browser action sequences |
get_metrics | current, history | Retrieve runtime telemetry, tick durations, and stuck events |
manage_behaviors | register, get, list | CRUD for immutable behavior tree JSON with SHA-256 tree hashes |
manage_blackboard | get, set | Read, write, or query shared behavior tree blackboard variables |
manage_runtime_db | stats, audit, snapshot, restore | SQLite diagnostics, SHA-256 Merkle audit, and checkpoint rollback |
🛡️ 5-Layer Safety Guardrail Stack & System 1 Invariants
- Fail-Closed Evaluator: Unrecognized node definitions throw fatal exceptions immediately.
- 60Hz Tick Invariant & Synchronous
semantic_check: The loop never blocks on external network calls; semantic condition checks resolve synchronously against blackboard caches. - HMAC Intention Gate Verification: Intentions dispatched to execution nodes require unexpired, cryptographically signed dispatch tokens (
PENTAD_HMAC_SECRET). - Action Rate Limiter & Leaf Policy Gate: Strict 60 actions/sec maximum throughput ceiling and policy guardrails blocking irreversible mutations.
- Emergency Kill Switch & Watchdog Heartbeat: Atomic safety latch and watchdog timers halting execution if tick stalls beyond 5,000ms.
📚 Deep Documentation Guides
- 📖 Formal API Reference: Full parameter tables, type definitions, and tool schemas.
- 💡 Core Architecture & Concepts: Behavior tree execution model, ~60Hz loop, and trigger lifecycle.
- 🖥️ CLI Usage Guide: Complete CLI command reference (
init,doctor,inspect,run). - 💾 Database Schema: SQLite table structures, indexes, and Merkle audit ledger.
- ⚙️ Configuration Reference:
.behavior-mcp.jsonparameters and environment variables.
🔗 Client Configuration & Environment
Add to .cursor/mcp.json or .vscode/mcp.json:
{
"mcpServers": {
"behavior-mcp": {
"command": "behavior-mcp",
"args": ["run"],
"env": {
"PENTAD_HMAC_SECRET": "your-secure-shared-secret-here"
}
}
}
}
🧪 Testing
# Run full unit and integration test suite across 38 test files (266 tests)
npm test
📄 License
MIT © PuterVision
Comments
New to MCP?
From what the protocol is to installing your first server, these answers cover most first questions.
An open protocol that defines how AI applications connect to external tools and data in a uniform way. Clients talk to the model; MCP servers provide tools, resources and prompt templates. They exchange standard messages, so one server can be reused by many clients.