Yaad Memory Fabric — Development skill for Claude Code
Yaad Memory Fabric — NVIDIA-powered knowledge/memory layer for the Open Agent Hackathon 2026 (Tinkerer Track).
How to install Yaad Memory Fabric
This entry records only its repository, not the path inside it, so there is no
exact command to give. Open devilking7x/yaad-memory-fabric and copy the folder into
~/.claude/skills/, or the file into ~/.claude/agents/.
What Yaad Memory Fabric does
Yaad Memory Fabric — NVIDIA-powered knowledge/memory layer for the Open Agent Hackathon 2026 (Tinkerer Track). Document ingestion, hybrid retrieval with citations, multi-agent research, and watch goals.
Alternatives in Development
- Create Handoff — You are tasked with writing a handoff document to hand off your work to another agent in a new session 10k ★
- MiniCode Architecture — 简体中文 This document describes the lightweight architecture decisions behind mini-code 388 ★
- AgentSkillOS — Build your agent from 200,000+ skills via skill RETRIEVAL & ORCHESTRATION 373 ★
README
Yaad Memory Fabric
**Open Agent Hackathon 2026 — Tinkerer Track entry.**
Yaad Memory Fabric is a new knowledge/memory layer that extends [Yaad](https://github.com/devilking7x/yaad) — Om's memory-based personal AI assistant — turning it from a chat assistant into a connected, autonomous knowledge agent, powered by NVIDIA sponsor technology.
🚧 Building in the open. Implementation lands during the official build window, **Oct 15–20, 2026** (judged work must be created during the window per the Tinkerer Track rules). This README tracks what exists vs. what is new — see `docs/NEW-WORK.md`.
What it adds to Yaad
| Module | What it does | Sponsor tech |
|---|---|---|
| Document ingestion | Upload PDF/MD/CSV/TXT → extract → chunk → embed → local vector store | NVIDIA NIM embeddings |
| Hybrid knowledge search | Vector + keyword retrieval with citations (doc + chunk) | NVIDIA NIM |
| Deep research v2 | planner → parallel investigators → critic → synthesizer, evidence-backed reports | NVIDIA NIM + Tavily |
| Watch goals | Standing goals → scheduled autonomous briefings | NVIDIA NIM |
Status
- Repo scaffolded (README, LICENSE, .gitignore)
- Implementation (build window Oct 15–20, 2026)
- Live demo deployment
- Demo video
Tinkerer story
Yaad already exists (agent loop, persistent memory, skills, web search). Memory Fabric is the **new work**: every file outside this seed commit is written during the build window and integrates NVIDIA NIM for embeddings, chat, and multi-agent reasoning over the user's own documents.
License
MIT — see [LICENSE](LICENSE).
Related Skills
Recallnest
One memory, every terminal. Shared memory layer for any MCP client — Claude Code, Codex, Kimi, Antigravity (ag
DSPy Multi Document Agents
An advanced distributed knowledge fabric for intelligent document processing, featuring multi-document agents,
Openshell Deepagent
A general-purpose coding agent that runs inside an NVIDIA OpenShell sandbox, orchestrated by Deep Agents and p
Flux Melange
Goal-seeking spice loop — adaptive review rounds steer toward the heat (novelty/risk/disagreement) found so fa
Embeddings Search Skill
/plugin install embeddings-search-skill@focus-marketplace Hybrid document search with semantic embeddings
Citation Validation
Verify all citations in a research document for accuracy, completeness, and quality.
Related Agents
Cozo Perf Architect
Use this agent when work touches CozoDB schema design, graph data modeling, indexes, query patterns, ingestion
Appsec AI Security Reviewer
Use for AI/LLM application security review: direct/indirect prompt injection, tool and agent authorization, RA
Python RAG Backend Engineer
Production-grade Python RAG backend engineer specializing in ChromaDB vector storage, sentence-transformers em