Python

Portable CLI for judging RAG and LLM benchmark runs across local, OpenAI-compatible, and cloud providers — a deterministic quick mode, a paraphrase-tolerant LLM-as-judge mode, and a full per-case audit trail for every verdict.

Apache AGE + pgvector in one PostgreSQL instance, side by side. Ask a question, watch vector-only, graph-only, and graph+vector retrieval run in parallel with timing for each.

Reads Claude Code’s own session transcripts and turns them into tokens, cost, time, and per-skill behavior — which prompt or skill is quietly ballooning your context, and which one keeps asking you questions.

GraphRAG that runs entirely in PostgreSQL — pgvector for vectors, recursive CTEs for graph traversal, tsvector BM25 for keyword search. No graph database, no second backup strategy, no data sync.

Sub-millisecond extractive summarization with byte-identical Python and Rust implementations. The preprocessor that sits in front of the LLM call and cuts tokens 40-94 percent.

Sub-millisecond extractive summarization with byte-identical Python and Rust implementations. The preprocessor that sits in front of the LLM call.

Standalone ingest-to-pgvector with a built-in chunker × embedder bakeoff. One YAML config = one end-to-end ingest cell. Python + Rust at parity.

Runnable sample app: Postgres 16 with Apache AGE and pgvector, a FastAPI orchestrator, and four retrieval strategies compared side by side on 391 real SCOTUS cases.

Scripts to test and benchmark JSON functionality across MySQL, PostgreSQL, and MongoDB — and generate simulated load.

Local-first MCP server that indexes code and docs into Postgres + pgvector for hybrid retrieval by LLM coding clients.