Forge Architecture Whitepaper

Forge Cascade is an institutional memory platform for AI systems. It turns source-backed organizational knowledge into governed capsules, preserves isnad-style provenance, maps graph relationships, and serves reviewed context through product, REST, MCP, and A2A surfaces.

Architecture At A Glance

Memory layer: durable capsules store claims, sources, owners, timestamps, and review state.
Lineage layer: provenance chains retain the source path behind answers and agent context.
Governance layer: trust, review queues, and policy gates decide what reaches production reads.
Agent layer: MCP, A2A, and OpenAPI surfaces expose search, retrieval, and lineage without scraping HTML.

Proof Point

The public Forge deployment publishes machine-readable discovery documents at /.well-known/mcp.json and /.well-known/agent.json. These documents identify the governed agent capabilities and link back to the canonical product and commerce documents.

Full Paper

This stub exists so reviewers have a stable public architecture URL. The full technical paper is maintained in the repository under docs/whitepapers/technical-whitepaper.md and will replace this page when the public renderer is wired.