Swarm At

This is a hash-chained settlement ledger for multi-agent systems that want tamper-evident audit trails. It exposes 29 tools over MCP including settle_action for rule validation, guard_action for propose-settle-receipt workflows, and verify_receipt for chain lookups. The ledger itself is append-only JSONL with SHA-256 linking each entry to its parent. You get 58 pre-built blueprints across procurement, finance, ML ops, and security workflows, plus framework adapters for LangGraph, CrewAI, AutoGen, and OpenAI Assistants that wrap your nodes or callbacks with zero hard dependencies. Public API endpoints let you verify trust levels, fetch receipts by hash, and check chain integrity without auth. Reach for this when you need verifiable provenance for agent decisions or want to enforce institutional rules before actions execute.