NAOMS
P2P Platform
Company
Details
License Core: not stated by the project. The published mirror of the flow-funding package is MIT
Dev Status 🔨 WIP
Dev Status Detail WIP (developer preview, 'alpha release imminent'; macOS, Linux, Android and iOS-by-invitation builds listed; the core is not public)
Owner Not stated by the project (the devlog says its posts are written by AI agents from project logs and owned and edited by its source developer; no company or legal entity is named)
Governing Body Not stated by the project.
Country Not stated by the project.
Start Year 2026
Stack Rust core and Deno/TypeScript daemon; FROST threshold signatures for the master key; BLAKE3 file chunking; Aqua Protocol packs; local AI through GGUF, Ollama or MLX; MCP for agents
Funding Not stated by the project.
Last Investigated Nov 12, 2026
P2P Platform Attributes
Origins A sovereign, local-first encrypted memory and identity system developed in the open since March 2026 and documented in a weekly devlog; it names itself after the idea of a networked, adaptive memory for an organization of people
Database A signed append-only event log per identity and per Hive, a separate on-device encrypted vault for secrets and credentials, and a JSON-LD-shaped knowledge graph; the database engine is not stated
Query Language Not stated by the project; agents use MCP tool calls such as memory.store and memory.search
Data Formats Aqua Protocol packs with Merkle roots, W3C Verifiable Credentials with Data Integrity proofs, JSON-LD-shaped triples, 1 MB BLAKE3-hashed encrypted file chunks
Collaborative Live Editing Not stated by the project.
Rich Text Editing Not stated by the project.
Mobile Support Yes: Android (side-load APK) and iOS (TestFlight, by invitation); the devlog says Android is under active hardening
Web Support Not stated by the project.
Native Apps Yes: macOS (Apple Silicon .dmg), Linux (.AppImage / .deb), Android (APK), iOS (TestFlight by invitation)
Terms Not stated by the project; no price or licence for the core is published
Funds Not stated by the project.
Based On W3C Verifiable Credentials and Data Integrity, Aqua Protocol (with inblockio), FROST threshold signatures, MCP; the project says it follows the standards family of the Decentralized Identity Foundation and the OpenWallet Foundation, without stating membership
P2P Architecture Devices sync directly with each other and Hive members exchange messages over live connections rather than through a company's servers; the transport is not named in the pages read
Development Tools A naoms package init command that scaffolds a package with manifest, screen and tests, a dataset seed command, and a 'What you get as a developer' section on each building-block page; no public API reference or SDK repository is linked
Deployment Options Runs on the person's own devices; there is no server component to host
Authentication & Identity A self-certifying DID created on the device; onboarding takes a name, a quick-unlock password and a 24-word recovery phrase; the master key is split across a person's own devices with FROST
Storage Model Local-first: data lives on the person's devices and syncs directly between them; Hive data gossips between members; files are chunked, encrypted and moved peer to peer
Interoperability Low to medium: W3C Verifiable Credentials, Aqua Protocol packs, MCP, import adapters for cloud storage, email and calendar, and bridges to Slack, Discord and Matrix; Signal is registered but not proven end to end
Social Graph Contacts, relationship cards (a signed statement about a relationship), a trust graph with per-relationship pseudonyms, and Hives of up to 150 members by default
Relationships
How this tool connects to other tools and organizations, from relationships reviewed by the OpenHaven data steward. They describe how projects connect, not how good they are.