citation-grounding

Developer acceleration layer for enterprise RAG + agent pipelines. 32 production-style flagships on LangChain, LangGraph, CrewAI, Temporal, OpenAI Agents SDK, pydantic-ai. Permission-aware retrieval, chunk-level citations, schema-enforced output. Built on Model Context Protocol (MCP).

Polish hub of legal-AI skills - 41 skills in 8 Claude Code / Cowork bundles (LLM output verification, PL/EU case law, DOCX redline, AI Act audit bundle). GDPR-safe, vendor-neutral, MIT.

Citation-grounded, refusal-bounded retrieval for regulated-domain corpora. Citation grounding and refusal-on-low-confidence enforced as deterministic checks in code, not prompt instructions. v0.1: FDCPA.

Traceable KG-RAG system for ideological education with citation grounding, multi-agent review, political safety checking, and cross-modal learning interaction.

A reproducible research study of citation-grounding and safety-routing for reducing hallucination in Taiwan-law legal assistants.

RAG over Indian law with verifiable citations: hybrid BM25 and vector retrieval fused by reciprocal rank fusion across roughly 19k chunks of judgments, statutes and case records.

Pre-call Intelligence for sales reps - turns a one-line meeting description into a skim-ready pre-call brief. It resolves the real companies and people involved, researches them across the web and court records, and produces a citation-grounded brief — no chat, no fluff.

Citation-grounded RAG assistant for semiconductor PDK/EDA engineers. Answers design-rule questions and drafts DRC/LVS snippet pseudocode, grounded in public PDK docs with a built-in citation guard.

Hybrid Search RAG over technical docs - pgvector + Postgres FTS fused via Reciprocal Rank Fusion, Cohere rerank-3.5, citation-grounded answers, Langfuse traces.

A production-grade MVP that challenges algorithmic bias and echo chambers in RAG systems. It audits global news and PDFs to uncover hidden risks and resilience in macro trends. Built with a Researcher–Auditor–Contrarian workflow using LangGraph, it runs locally with Ollama and OpenSearch for privacy.

End-to-end governed document intelligence platform with citation-grounded RAG, schema-constrained extraction, human review workflows, and immutable audit trails.

Cited runbook retrieval with quality history, answer-safety gates, CLI/API parity, Prometheus, Docker, and Kubernetes.

Local first RAG assistant for consulting delivery: grounded Q&A and structured ADR/story/risk outputs with citations, evals, and observability

English hub of method-neutral legal AI skills (verification core, content quality, EU law) - bundle model. Polish-jurisdiction skills live in awesome-matematic-skills-pl.