Your Personal AI super intelligence. Private, Simple and extremely powerful.
| Rating | Summary | |
|---|---|---|
| Quality | excellent (21/24) | Rapidly iterated, heavily adopted Rust/Tauri desktop personal AI with strong packaging discipline, active maintenance, and solid multi-language documentation. |
| PAI Relevance | watch (0.25) | High-quality personal AI product but an unintegrable GUI app with near-zero composite fit — worth monitoring as the category leader, not incorporating. |
Formula gives WATCH (standalone 21 ≥ 12); composite 0.25 is well below the INTEGRATE threshold and integration score is 0 (Tauri GUI, no subprocess or SDK surface exposed). Prior landscape entry recorded this as SKIP; formula WATCH is retained as the correct first-principles result given the high standalone quality.
21/24 — actively-maintained / adequately-documented / high-discipline
Failed:
Passed:
Failed:
Passed:
Failed:
Passed:
| Dimension | Score | Assessment |
|---|---|---|
| Harvest Value | 1 | The Memory Tree + Obsidian-style vault pattern and Composio integration layer offer mild design interest, but PAI already has analogous capabilities in Knowledge (persistent knowledge graph), ContextSearch (session recovery), and Research (web + vault investigation). No novel algorithm or architecture is directly applicable to a specific PAI subsystem. |
| Integration Readiness | 0 | OpenHuman is a Tauri GUI desktop application; it cannot be bun add-ed or cleanly subprocessed into PAI's CLI-first workflow. No API, SDK, or stable IPC surface is exposed. This is a product, not a library or CLI. |
| Overlap Risk | 1 | Partial overlap with PAI's Knowledge (persistent knowledge graph), Research (web investigation + vault), and Agents (agent composition + voices) skills. Operates at the consumer application layer rather than PAI's infrastructure layer, so overlap is functional but not structural. |
| Gap Fill | 0 | PAI's existing capabilities (Knowledge, Research, ContextSearch, Agents orchestration) already cover the functional space OpenHuman inhabits. The desktop GUI layer is orthogonal to PAI's CLI-first, Bun-based design. |
Composite: 0.25
Local-first agent memory architecture: Install the current v0.56.0 release, connect a model, and inspect the on-disk Obsidian vault structure that OpenHuman writes as its Memory Tree — map the folder layout and note format before the schema changes again. If the format stabilizes over the next two releases, it becomes a concrete reference design for portable, human-readable agent memory that requires no vector database and no cloud dependency.
Non-technical AI tool distribution UX: Walk through the full first-run onboarding as a non-technical user would — model connection, first prompt, vault creation — and document where friction still occurs. OpenHuman's 29k-star growth in three months is evidence this distribution pattern (native GUI, BYO API key, zero terminal) is solving a real problem; the friction points you find are directly actionable if you ever ship a local-first AI tool to non-developer users.
Composio-backed tool-calling layer: Track whether OpenHuman's Composio integration contract gets extracted into a documented spec or standalone SDK — it's currently coupled tightly to the GUI with no independent interface. Revisit in 3 months: if the tool-calling layer ships as a separable component, it becomes a candidate for wiring local model endpoints to external services without writing custom connectors; if it stays embedded, the repo remains a reference implementation only.
Category: Personal AI & Knowledge
In this category: First and only formal taxonomy entry; overlaps with uncategorized personal AI cluster containing vellum-ai--vellum-assistant, garrytan--gbrain, NorthwoodsSentinel--loam, MemPalace--mempalace, malteristo--turtleos, and UnluckyMycologist68--palimpsest.
Standing: Dominant by every metric — 29k stars and 2.7k forks against single-to-low-double-digit hundreds for nearest personal AI peers; the only entry with GPG-signed native package distribution across all three OS platforms and a versioned release cadence.
Density: 9/10 — Available: full README (8KB), root package.json dependency manifest, repo metadata (stars, forks, issues, license, dates), and release history (v0.56.0). Missing: workspace-level package.json(s) with actual app dependency counts, Cargo.toml, CI workflow files, and full codebase structure.