Behavior Runtime Engine (behavior-mcp)
SkillWeb & browsingTeaches the agent to use the Behavior MCP server for ~60Hz in-browser behavior trees, triggers, and recordings.
Available today. Use it from your connected AI after setup.
No other account needed.
Add ahel to your AI once: Claude, ChatGPT, Cursor, Claude Code or Codex. Then ask it to use this.
Then ask your AI: use the Behavior Runtime Engine (behavior-mcp) skill
What this skill tells your AI
The instructions your AI receives, as published by putervision/world-model-mcp in .agents/skills/behavior-mcp/SKILL.md and read by ahel’s review.
This skill provides step-by-step guidance and operational patterns for interacting with @putervision/behavior-mcp with project slug "world-model-mcp".
1. Role in the PuterVision Pentad
- Workflow State (
state-memory-mcp): Persistent task DAGs, decisions, milestones, and blockers. - Perception (
vision-memory-mcp): Visual layout caching, screenshots, and visual specifications. - Spatial World (
world-model-mcp): Persistent 3D/2D coordinates, bounding boxes, and topological relations. - Strategic Reasoning (
agent-reasoning-mcp): Strategic BDI goals, utility scoring, and intention dispatch. - Tactical Execution (
behavior-mcp): High-frequency (~60Hz) deterministic behavior tree execution directly in browser runtimes with reactive interrupts.
2. Core Operational Sequence
- Define or Load Behavior: Call
load_behaviorwithaction: "load"to inject and start tree execution. - Register Reactive Triggers: Configure emergency preemption using
register_triggerwith priority and cooldown guards. - Monitor Telemetry: Poll execution health, tick rates, and active traversal nodes with
get_status. - Runtime Blackboard: Read and mutate execution variables via
manage_blackboard. - Safety Guardrails: Immediately halt or pause rogue states using
abort_behavior.
3. Complete 10 Consolidated MCP Tools Reference
| Tool Name | Key Actions | Key Parameters | Description |
|---|---|---|---|
load_behavior | load, unload, swap | behavior_name, behavior_version, parameters, intention_id | Hot-swap, inject, or initialize behavior tree instances in browser runtime. |
set_parameters | set, get, reset | execution_id, parameters | Dynamically update or query execution parameters for the active tree. |
get_status | current, history, tree_state | execution_id, limit | Query active behavior status, current node path, tick count, duration, and errors. |
abort_behavior | abort, pause, resume | execution_id, reason | Immediately halt, pause, or resume execution and disengage active inputs. |
register_trigger | register, list | name, condition_type, priority, cooldown_ms, behavior_name | Configure reactive interrupt triggers with priority preemption and cooldowns. |
replay_recording | capture, list | recording_id, execution_id, frames, name | Capture or replay deterministic browser action sequences with adaptive timing. |
get_metrics | current, history, aggregate, compare | execution_id, behavior_name, limit | Retrieve execution telemetry, tick durations, stuck events, and statistics. |
manage_behaviors | register, list, get | name, version, tree, tree_json, description | CRUD operations for immutable JSON behavior trees with SHA-256 tree hash verification. |
manage_blackboard | get, set | execution_id, key, value | Read or write shared behavior tree blackboard state variables. |
manage_runtime_db | stats, audit, snapshot, diff, restore | action, name, description | Database maintenance, diagnostics, and SHA-256 Merkle audit verification. |
Signals
- GitHub stars
- 46
- Forks
- 3
- Last commit
- Sep 2026
Others that do the same job
Advanced
- Item type
- skill
- Key
behavior-mcp-putervision- Source
- github.com/putervision/world-model-mcp