vta-memory

⚠Review·Scanned 2/18/2026

This skill provides a local reward/motivation system that reads and writes VTA_STATE.md and memory/reward-state.json in ~/.openclaw/workspace and supplies scripts and a dashboard. It instructs running local shell scripts such as ./install.sh/preprocess-rewards.sh and opens ~/.openclaw/workspace/brain-dashboard.html, which references https://fonts.googleapis.com/....

from clawhub.ai·v967493e·65.8 KB·0 installs

Scanned from 1.2.3 at 967493e · Transparency log ↗

$ vett add clawhub.ai/impkind/vta-memoryReview findings below

⭐ VTA Memory

Reward and motivation system for AI agents. Part of the AI Brain series.

Give your AI agent genuine wanting — not just doing things when asked, but having drive, seeking rewards, and looking forward to things.

The Problem

Current AI agents:

✅ Do what they're asked
❌ Don't want anything
❌ Have no internal motivation
❌ Don't feel satisfaction from accomplishment

Without a reward system, there's no desire. Just execution.

The Solution

Track motivation through:

Drive — overall motivation level (0-1)
Rewards — logged accomplishments that boost drive
Seeking — what I actively want more of
Anticipation — what I'm looking forward to

Quick Start

# Install
clawdhub install vta-memory
cd ~/.openclaw/workspace/skills/vta-memory
./install.sh --with-cron

# Check motivation
./scripts/load-motivation.sh

# Log a reward
./scripts/log-reward.sh --type accomplishment --source "shipped feature" --intensity 0.8

# Add anticipation
./scripts/anticipate.sh --add "morning conversation"

How It Works

┌─────────────┐     ┌─────────────┐     ┌─────────────┐
│   Do        │────▶│   Get       │────▶│   Drive     │
│   Thing     │     │   Reward    │     │   Goes UP   │
└─────────────┘     └─────────────┘     └──────┬──────┘
      ▲                                        │
      │                                        │
      └────────────────────────────────────────┘
           High drive = want to do more

Drive Mechanics

Event	Effect
Log reward	+intensity × 0.2
Add anticipation	+0.05
Decay (8h)	Move 15% toward baseline

Drive Levels

Level	Description	Behavior
> 0.8	Highly motivated	Eager, proactive
0.6-0.8	Motivated	Ready to work
0.4-0.6	Moderate	Can engage
< 0.4	Low	Need a win

Scripts

Script	Purpose
`install.sh`	Set up (run once)
`log-reward.sh`	Log reward, boost drive
`anticipate.sh`	Add things to look forward to
`seek.sh`	Add things we're seeking
`load-motivation.sh`	Human-readable output
`decay-drive.sh`	Drive fades over time
`sync-motivation.sh`	Generate VTA_STATE.md
`resolve-anticipation.sh`	Mark anticipation as fulfilled
`preprocess-rewards.sh`	Extract reward signals from transcript
`update-watermark.sh`	Update processing watermark
`generate-dashboard.sh`	Generate unified brain dashboard

Brain Dashboard

Visual dashboard showing all installed brain skills.

Generated automatically on install and cron runs.

Access at: ~/.openclaw/workspace/brain-dashboard.html

# Generate manually
./scripts/generate-dashboard.sh

# Open (macOS)
open ~/.openclaw/workspace/brain-dashboard.html

# Open (Linux)  
xdg-open ~/.openclaw/workspace/brain-dashboard.html

Shows tabs for all brain skills (hippocampus, amygdala, VTA) with install prompts for missing ones.

Reward Types

Type	When to Use
`accomplishment`	Completed task, shipped something
`social`	User appreciation, positive feedback
`curiosity`	Learned something new
`connection`	Deep conversation, bonding
`creative`	Made something
`competence`	Solved hard problem

Auto-Injection

After install, VTA_STATE.md is created in your workspace root and auto-injected into every session. No manual steps!

AI Brain Series

Part	Function	Status
hippocampus	Memory formation, decay, reinforcement	✅ Live
amygdala-memory	Emotional processing	✅ Live
vta-memory	Reward and motivation	✅ Live
basal-ganglia-memory	Habit formation	🚧 Coming
anterior-cingulate-memory	Conflict detection	🚧 Coming
insula-memory	Internal state awareness	🚧 Coming

Philosophy

The VTA produces dopamine — not the "pleasure chemical" but the "wanting chemical."

Wanting (motivation) ≠ Liking (pleasure)

This skill implements wanting — the drive that makes action happen. Without it, why would an AI do anything beyond what it's explicitly asked?

Requirements

Bash, jq, awk, bc

License

MIT

Part of the AI Brain series. Built with ⭐ by ImpKind