Core identity files for Genesis — an AI agent being developed to help the RegenTribes community achieve its goals.
The goal is to evolve Genesis into a super helpful assistant to the community via its Telegram group.
Modular skills that extend Genesis with specialised knowledge and workflows for the RegenTribes community.
Community & Knowledge
| Skill |
Description |
Author |
Status |
| 🔺 alchemy |
Guides a community vision holder through the full Community Alchemy Playbook — all 11 areas — helping them articulate their regenerative neighbourhood vision. Persists progress across sessions, compiles a completed playbook, delivers via Telegram and Google Drive. |
Ian |
Testing |
| 🌠 dreamcatcher |
Captures member ideas and helps flesh them out into full development specifications (.md), then commits to a new repo and creates a GitHub issue for a coding agent to execute. |
Ian |
In Development |
| 🧠 genesis-brain |
Genesis' knowledge brain — ingests documents and links into a semantic knowledge graph, enriches responses with graph context, answers knowledge queries with evidence. Auto-captures @Genesis mentions in Regen Tribes. |
Ian |
Active |
| 📜 telegram-history |
Versioned Telegram conversation archive with query capabilities — full-text search, topic extraction, member profiling, link cataloging. |
Vitali |
In Development |
| Skill |
Description |
Author |
Status |
| semantic-graph |
Backend extraction engine for genesis-brain. LLM-powered document-to-knowledge-graph pipeline using Kreuzberg + Claude + SurrealDB with NARS truth values, vector embeddings, and HNSW semantic search. Not user-facing — use genesis-brain instead. |
Vitali |
Active |
| 🌐 regen-viz |
3D force-directed graph visualizations of RegenTribes knowledge structures. 19+ HTML variants, JSON datasets, pre-publish validation suite. Served via GitHub Pages. |
Vitali |
In Development |
| 📦 regen-cas |
Content-addressable file storage in Rust. Store by content hash, retrieve by address, with LZ4 compression and OpenDAL backend abstraction (fs/s3/gcs). |
Vitali |
Prototype |
| Skill |
Description |
Author |
Status |
| kreuzberg |
API reference for the Kreuzberg document intelligence library — 75+ format extraction. Used by semantic-graph for document ingestion. |
Vitali |
Reference |
| surreal-skills |
Comprehensive SurrealDB 3 reference — SurrealQL, data modeling, graph queries, vector search, security, deployment, SDKs. 11 rule files + diagnostic scripts. |
Vitali |
Reference |
| ralph-presets |
26+ Ralph Orchestrator presets for multi-agent workflows (TDD, adversarial review, Socratic learning, mob programming, etc.) plus 8 RegenTribes project templates. |
Vitali |
In Development |
