Show HN: Agenda Intel MD – schemas and CLI to audit LLM strategic-risk briefs
Agenda Intelligence MD is an open-source toolkit designed to audit and validate the structure of AI-generated strategic-risk briefs, ensuring they meet rigorous evidence and reasoning standards. It provides a protocol, JSON schemas, CLI tools, and an MCP server to help AI agents produce auditable, evidence-backed outputs rather than plausible-sounding summaries. The toolkit targets engineers building policy, geopolitical risk, and regulatory intelligence agents that must withstand analyst review.
- ▪Agenda Intelligence MD includes a markdown protocol, JSON schemas, CLI tools, and an MCP server for validating and scoring strategic-risk briefs.
- ▪It ensures AI outputs are structurally sound, evidence-labeled, and decision-focused, but does not verify factual accuracy.
- ▪The toolkit supports domains like sanctions, export controls, and regional analysis in Central Asia, the Gulf, and the Middle East.
- ▪It comes with a benchmark set of five cases, including the EU AI Act and Red Sea shipping, achieving 100% schema validity and evidence packing.
- ▪Agenda Intelligence MD is not a factuality checker, source retriever, or autonomous news agent, and does not replace human analyst judgment.
Opening excerpt (first ~120 words) tap to expand
Agenda Intelligence MD CI / MCP / EVIDENCE-AUDIT LAYER FOR STRATEGIC INTELLIGENCE AGENTS — protocol, JSON schemas, CLI and MCP server for validating, scoring and auditing the structure of strategic-risk agent output. The evidence-discipline surface for markdown-first reasoning skills (Global Think Tank Analyst, Central Asia + Caspian, Gulf + Middle East). Open-source. Evidence & eval layer for strategic intelligence agents. A protocol, JSON-schema set, CLI, and MCP-compatible toolkit that helps AI agents move from unsupported summaries to auditable strategic-risk briefs: what changed why it matters what is evidence-backed what is uncertain who gains or loses leverage what scenarios are plausible what to watch next It is built for engineers shipping policy, sanctions, regulation,…
Excerpt limited to ~120 words for fair-use compliance. The full article is at GitHub.