Pinned Loading

1. prompt_compressor prompt_compressor Public

   Compresses LLM prompts while preserving semantic meaning to reduce token usage and cost.

   Python 7 1

2. medbert2rag_router medbert2rag_router Public

   NLU router from natural language query to SPARQL, RAG and MedBert LLM model.

   Python

3. nergrep nergrep Public

   A Python/NLP package and CLI for extracting and filtering named entities using spaCy, with powerful filtering and seamless integration into modern LLM pipelines.

   Python

4. puppetry-detector puppetry-detector Public

   **Puppetry Detector** is a modular engine for detecting structured and semantic prompt injection attacks against Large Language Models (LLMs).

   Python 1

5. multi-chain-balance-diff multi-chain-balance-diff Public

   Diff wallet balances across 13+ chains (Ethereum, Polygon, Base, Arbitrum, Solana, TON, etc). JSON output, threshold alerts, webhooks, CI-friendly exit codes. A scriptable primitive for treasury mo…

   JavaScript

6. tiny-rag-core tiny-rag-core Public

   RAG in 8 lines of shell. Normalize → score → rank → return. A contemplative miniature.

   Shell