Setup & Installation
openclaw skills install @fatcatmaofei/mindcore
Or with the ClawHub CLI, for registry-managed skill folders outside a full OpenClaw workspace:
npx clawhub install mindcore
What This Skill Does
MindCore is a background daemon that simulates a subconscious mind for AI agents. It runs a 5-layer neural pipeline to generate spontaneous impulses like hunger, boredom, or social urges, outputting JSON signals when a thought crosses its firing threshold. Runs locally on CPU with no LLM dependency.
Because it runs entirely on CPU without an LLM, it adds emotional state to agents without extra inference cost or latency.
When to use it
- Adding unprompted check-ins to a companion AI agent
- Simulating fatigue and hunger cycles in a virtual character
- Triggering spontaneous topic changes in a long-running chat agent
- Driving mood-influenced response variation throughout the day
- Prototyping emotionally reactive NPCs in text-based environments
Example Workflow
Here's how your AI assistant might use this skill in practice.
INPUT
User wants their AI agent to occasionally initiate conversation on its own
AGENT
- 1Start MindCore daemon with python main.py
- 2Agent listens on MindCore's JSON output stream
- 3A social impulse fires after accumulating past threshold
- 4Agent receives signal: {"impulse": "I suddenly want to chat", "strength": 0.83}
- 5Agent uses signal to trigger an unprompted message to the user
OUTPUT
Agent sends: "Hey, I've been thinking about something — do you want to talk?"