genai regulatory intel.stub

---

id: genai-regulatory-intel name: RegIntel-AgenticAI — GenAI Regulatory Intelligence repo_path: ~/Documents/sw30labs/repos/genai-regulatory-intel remote_url: https://github.com/sw30labs/genai-regulatory-intel.git primary_language: Python framework: LangGraph + OpenAI Responses API category: oscal-compliance stacks: [agentic, langgraph, compliance, rag] ingested: 2026-05-29 last_commit_date: 2026-05-25

Purpose

Autonomous regulatory intelligence CLI for tracking, extracting, and comparing GenAI regulations across 16 global financial services jurisdictions (EU, UK, USA, Singapore, HK, China, Australia, Switzerland, and more). Built for Tier 1 banks navigating the fast-evolving GenAI regulatory landscape — EU AI Act, US state-level laws, APAC frameworks.

Architecture

• 4-Agent State Machine: Research → Validate → Reflect → Persist (LangGraph graph)
• Web Search: OpenAI Responses API hosted search for current regulatory sources
• Hybrid Storage: SQLite (structured regulatory data) + ChromaDB (semantic search)
• Jurisdictions config: jurisdictions.yaml drives the 16 target regions
• Gap Analysis: Identifies uncovered regulatory themes across jurisdictions
• Comparison Matrices: Side-by-side GenAI regulatory requirements
• Token tracking: Input, output, reasoning, and cached token totals per run

Dependencies

• LangGraph (multi-agent orchestration)
• OpenAI Responses API (research agents + web search)
• ChromaDB (semantic retrieval)
• SQLite (persistent structured storage)
• pyproject.toml / Python 3.10+

CI/CD

42 tests (pytest).

Entities and topics

• oscal-compliance
• agentic-frameworks
• Tags: [langgraph, compliance, rag, llm, nist, security]