Your Personal AI super intelligence. Private, Simple and extremely powerful.
INTEGRATE — Pull into vault as the canonical personal AI desktop harness reference; Memory Tree hierarchy, TokenJuice compression pipeline, and auto-fetch loop are directly harvestable architecture patterns.
20/24 — actively-maintained / adequately-documented / high-discipline
Failed: H8: FAIL — README contains only status and release badges (shields.io); no GitHub Actions CI badge and no explicit reference to .github/workflows/ found in the provided content
Passed: H1: PASS — release v0.53.22 tagged 2026-05-09 H2: PASS — release 2026-05-09 is 4 days ago, well within 12 months H3: PASS — last commit 2026-05-13 (today), within 6 months H4: PASS — last commit 2026-05-13, within 30 days H5: PASS — archived: false H6: PASS — 58 open issues, within healthy range (>0 and <100) H7: PASS — GPL-3.0 license declared
Failed: D4: FAIL — code blocks for installation appear before any heading; no heading containing "usage", "example", "quickstart", or "getting started" precedes a code block in the README D5: FAIL — no heading with "API", "Configuration", "Options", "Reference", "Commands", or "Parameters" found in README; configuration detail deferred to external gitbook docs D8: FAIL — early-beta caveat is inline prose ("> Early Beta: Under active development. Expect rough edges."), not a dedicated section heading for Limitations, Caveats, or Known Issues
Passed: D1: PASS — README is present and non-empty D2: PASS — README substantially exceeds 1000 bytes; multiple long feature sections, comparison table, install instructions D3: PASS — README contains curl install script for macOS/Linux and PowerShell script for Windows, plus link to download page D6: PASS — first 500 characters include "OpenHuman is your Personal AI super intelligence. Private, Simple and extremely powerful." D7: PASS — README links to https://tinyhumans.gitbook.io/openhuman/ (external docs) and multiple sub-docs pages
Failed: none
Passed: E1: PASS — primary language is Rust (Tauri backend), TypeScript frontend; both typed E2: PASS — package.json present with pnpm workspace configuration E3: PASS — package.json shows 3 direct dependencies (@tauri-apps/api, husky, ws), well under 30 E4: PASS — package.json includes "test", "test:coverage", "test:rust" scripts; husky pre-commit hooks present E5: PASS — 2988 stars, far exceeds 50 E6: PASS — created 2026-02-18, ~2.8 months ago; ~1079 stars/month, far exceeds threshold of 2 E7: PASS — 291 forks, far exceeds 5 E8: PASS — description "Your Personal AI super intelligence. Private, Simple and extremely powerful." is meaningful, >20 chars, not just repo name
| Dimension | Score | Assessment |
|---|---|---|
| Harvest Value | 2 | Memory Tree hierarchical summarization into ≤3k-token Markdown chunks, the TokenJuice compression pipeline (HTML→Markdown, URL shortening, non-ASCII strip), and the 20-minute auto-fetch loop pattern are novel, directly applicable approaches to PAI's memory and context-window management concerns. |
| Integration Readiness | 0 | OpenHuman is a monolithic Tauri desktop application (Rust + TypeScript); extracting individual subsystems (Memory Tree, TokenJuice, integration connectors) for use inside PAI would require more than a day of adaptation work and is constrained further by GPL-3.0 copyleft. |
| Overlap Risk | 1 | Partial overlap with garrytan--gbrain (personal memory vault, knowledge graph, hybrid RAG) and NorthwoodsSentinel--loam (personal AI memory substrate); however, OpenHuman operates at a higher abstraction level as a full desktop app with integrations and voice, covering different surface area than those substrate libraries. |
| Gap Fill | 2 | The landscape Gaps section explicitly flags "Personal AI Memory: needs alternatives for comparison"; OpenHuman is a fully-realized, high-star alternative that adds depth to that thin category and directly demonstrates production-grade approaches to the memory and context problems flagged there. |
Composite: 0.63
Category: Personal AI Desktop Agent Crowding: 0 repos in vault (first-in-category) Alternatives: garrytan--gbrain and NorthwoodsSentinel--loam are in the related Personal AI Memory category; no prior desktop agent harness appraised vs. top alternative: Where garrytan--gbrain is a TypeScript memory substrate requiring BYO agent harness, OpenHuman is a batteries-included desktop application with 118+ OAuth integrations, voice, mascot, and model routing already wired together Landscape impact: filling a gap — first full-stack personal AI desktop agent appraised; adds a production reference tier above the substrate-level memory repos already in the vault
Density: 9/10 — Available: README (full, ~8KB), repository metadata (stars/forks/issues/dates), dependency manifest (package.json with workspace scripts), license (GPL-3.0), release history (v0.53.22), language detection (Rust primary), description, commit timestamps, archived status. Missing: full CI/workflow configuration files (.github/workflows/ contents not provided).
GPL-3.0 licensing is a meaningful constraint for any direct code reuse — harvest should focus on architectural patterns and design decisions rather than copy-paste. The v0.53.22 release number (53 minor versions in ~3 months of existence) implies extremely rapid release cadence, which is consistent with the "early beta" label; stability of APIs and internal architecture should be assumed volatile. The Trendshift badge and ~1079 stars/month growth rate suggest this repo will accumulate significant community tooling and plugin ecosystem quickly — worth monitoring for emerging integration patterns (MCP endpoints, external API surface) that could lower integration readiness in future reappraisals.