# Halofy

An open, self-hostable access and governance layer that gives every AI agent in an organization the same identity, policy and audit boundary over MCP or HTTP, with signed, independently verifiable erasure.

Halofy sits between organizational knowledge and the agents that use it, resolving identity, enforcing role- and namespace-scoped access, governing change and recording every outcome so agents, applications and human operators share one server-resolved identity and policy boundary instead of each building its own. It ships one governed access surface, mem_write, mem_read, mem_search, mem_assemble, mem_fault, mem_stats, mem_forget, plus policy, sharing, pinning, export and manifest operations, over both MCP and HTTP.

Underneath that boundary it maintains a governed context lifecycle with provenance, entity resolution, exact and semantic dedupe, supersedence chains and append-only audit, storing operational context in Postgres/pgvector, with an embedded PGlite zero-setup default, and cold-tier knowledge as encrypted, Git-versioned data. Four built-in connectors, filesystem, Postgres, Obsidian and manual CSV import, all feed the same governed write path, and a hermetic test lane runs entirely offline with a stub LLM and a hash-based embedder.

Website: https://halofy.ai