vibeship-spark-intelligence 0.1.1

Self-evolving intelligence layer for AI agents

Spark Intelligence

self-evolving intelligent companion.

---

Learns constantly. Resonates and evolves with you. Runs 100% on your machine as a local AI companion. Brings a spark to everything.

You code → Spark learns → Agent adapts → You code better → Spark learns more

What is Spark?

Spark Intelligence is a self-evolution layer for AI coding agents. It captures interaction signals, distills them into learnings, and feeds those learnings back into your agent to improve behavior over time.

Not a chatbot. Not a wrapper. A learning engine.

Install

git clone https://github.com/vibeforge1111/vibeship-spark-intelligence
cd vibeship-spark-intelligence
pip install -e .[services]

Quick Start

# Start services
spark up

# Check health
spark health

# View what Spark has learned
spark learnings

Windows: run start_spark.bat from the repo root.

Lightweight mode (core only, no dashboards): spark up --lite

Connect Your Agent

Spark works with any coding agent that supports hooks or event capture.

Agent	Integration	Guide
Claude Code	Hooks (PreToolUse, PostToolUse, UserPromptSubmit)	docs/claude_code.md
Cursor / VS Code	tasks.json + emit_event	docs/cursor.md
OpenClaw	Session JSONL tailer	docs/openclaw/

What You Get

• Learning engine — captures signals, distills insights, promotes high-value learnings to your agent context
• Quality gates — Meta-Ralph scores every insight before it enters the knowledge base
• Advisory system — pre-tool advice ranked by fusion scoring across 7 sources
• Episodic intelligence (EIDOS) — prediction → outcome → evaluation loop
• Domain chips — pluggable YAML modules for domain-specific learning
• Dashboards — Spark Pulse (primary), Meta-Ralph analyzer
• CLI — spark status, spark learnings, spark promote, spark up/down, and more
• Hot-reloadable config — tuneables system with schema validation and drift tracking

Architecture

Your Agent (Claude Code / Cursor / OpenClaw)
  → hooks capture events
  → queue → bridge worker → pipeline
  → quality gate (Meta-Ralph) → cognitive learner
  → advisory system delivers insights pre-tool
  → context files updated → agent reads and adapts

Documentation

• 5-minute start: docs/GETTING_STARTED_5_MIN.md
• Full setup: docs/QUICKSTART.md
• Docs index: docs/DOCS_INDEX.md
• Website: spark.vibeship.co
• Contributing: CONTRIBUTING.md (local setup, PR flow, and safety expectations)

Responsible Use

This is a self-evolving system. If you are planning a public release or high-autonomy deployment:

• Read first: docs/AI_MANIFESTO.md
• Read first: https://aimanifesto.vibeship.co/
• docs/RESPONSIBLE_PUBLIC_RELEASE.md
• docs/security/THREAT_MODEL.md
• SECURITY.md for vulnerability reporting

License

MIT — free to use, modify, and distribute.

---

_{Built by Vibeship}