Read 2026-10-23
tinyhumansai/openhuman: technical due diligence, read from its repository
What Borehole read in tinyhumansai/openhuman's code and history at adac27a: its licence, security advisories, releases, review practice, tests, dependencies and operations. Every finding cites the file and line it rests on. The full report also covers the team and asks the questions the code cannot answer.
| Dimension | Rating |
|---|---|
| Architecture & Codebase — including AI provenance | strong |
| Engineering Process & SDLC Maturity | strong |
| Product & Engineering Maturity | not assessable |
| Security & Reliability Posture | adequate |
| Observability & Operations | not assessable |
| Fit with the Acquirer | not assessable |
What the repository says
.gitmodules:1: 21 submodule(s) this survey does not read: vendor/tinyagents, vendor/tinyflows, vendor/tinychannels
Architecture & Codebase — including AI provenance
strongapp/src/lib/i18n/ja.ts is 5,459 lines, 39× this codebase's median file
adequateThe median source file here is 138 lines. 27 files exceed 1,380 lines, the largest being 5,459. Size is not a defect on its own — some problems genuinely live in one place — but these are where merge conflicts, review fatigue and single-owner knowledge concentrate, and they are the first thing that slows an inheriting engineer down.
app/src/lib/i18n/ja.ts:1 — 5,459 lines (39× median) |
app/src/lib/i18n/en.ts:1 — 5,380 lines (38× median) |
app/src/lib/i18n/de.ts:1 — 5,253 lines (38× median) |
app/src/lib/i18n/fr.ts:1 — 5,228 lines (37× median) |
app/src/lib/i18n/es.ts:1 — 5,206 lines (37× median) |
Engineering Process & SDLC Maturity
strongTests are retried until they pass
adequateThe suite is configured to retry failing tests (app/playwright.config.ts). A retry turns a flaky test green without saying why it failed, so a green build means the tests passed eventually. Ask how many retries a typical run needs, and which tests need them.
app/playwright.config.ts:9 — test retry |
Product & Engineering Maturity
not assessableNot rated: only 7 of 19 checks could reach a conclusion here; the rest do not apply to this repository or could not read it.
Security & Reliability Posture
adequate1 credential-shaped string in the tree sit in tests, examples, documentation or local development settings
adequateStrings shaped like credentials appear only where fakes are usual: test fixtures, examples, documentation, placeholders and local stacks. They are almost always fakes. Ask someone to confirm that none of them works anywhere, because a real key in a development file is still a real key.
| GitHub token, where fakes are usual |
2 Dockerfiles run the application as root
adequateNo USER instruction switches away from root, so the process runs with full privileges inside the container. It is a one-line fix and an ordinary expectation, which is why its absence is worth asking about: it usually says more about whether anyone has reviewed the deployment than about this container specifically.
.oss-scanner/Dockerfile — runs as root |
Dockerfile — runs as root |
The build runs third-party code it does not pin
adequate13 third-party CI actions are referenced by a tag or branch rather than a commit, so they run whatever that tag points to on the day; 7 steps pipe a download straight into a shell. Tags have been moved to malicious commits in real supply-chain attacks, and a piped script is whatever the server returns that minute. Pinning to a commit is the fix, and it is small.
.github/workflows/android-compile.yml:101 — action pinned by tag |
.github/workflows/android-compile.yml:107 — action pinned by tag |
.github/Dockerfile:34 — download piped into a shell |
54 locked dependencies have published vulnerabilities, 13 high or critical in what ships
adequate54 of 2,849 locked dependencies match 158 published advisories in OSV, 67 of them rated high or critical. A lockfile says which version is installed, not whether the flawed code is reachable from this product, so each needs a look. Most are fixed by an upgrade, and the advisory names the version. This lockfile pins development tools alongside what ships and does not say which is which, so some of these may never reach a user.
pnpm-lock.yaml — @xmldom/xmldom 0.8.13: GHSA-27p8-2357-5qqv (high), fixed in 0.8.15 |
pnpm-lock.yaml — react-router 7.14.2: GHSA-8x6r-g9mw-2r78 (high), fixed in 7.15.0 |
pnpm-lock.yaml — socket.io-parser 4.2.6: GHSA-2m8v-j782-fhvr (high), fixed in 4.2.7 |
The production build publishes source maps
adequateThe build configuration emits source maps for production, so the original source, with its comments and structure, is downloadable by every visitor. Upload them to an error tracker and keep them off the public server instead.
app/vite.config.ts:281 — source maps in a production build |
12 advisories against this project in the last 24 months, 10 high or critical
adequateThe newest was published on 2026-10-09. A project that publishes advisories is telling its users about its fixes, and the count and pace say how much security work it carries. Whether 12 of them are fixed in the code you ship could not be told from this history: its version is only known to be at least 0.64.15. Answer lives in: the release you run, set against each fixed version.
| 2026-10-09: GHSA-g6j9-89xv-v752 (high), fixed in >= 0.57.13: MCP tool description injection: untrusted server descriptions execute in agent LLM context without sanitization |
| 2026-10-09: GHSA-p9gw-q7cf-57qr (high), fixed in >= 0.57.13: Indirect prompt injection via background email sync executes Composio actions without user approval |
| 2026-10-09: GHSA-gg3m-wgmq-vhmx (critical), fixed in >= 0.57.13: Yuanbao non-web channel can bypass human approval to execute `node_exec` |
| 2026-10-09: GHSA-f46p-6vf9-64mm (critical), fixed in >= 0.57.13: OpenHuman allows remote inbound channel messages to create unapproved shell cron jobs leading to host command execution |
| 2026-10-09: GHSA-ww8p-c82p-p8wx (high), fixed in >= 0.57.5: Bearer token leaked via approval gate session_id (SQLite plaintext + debug log + RPC response) |
Observability & Operations
not assessableNot rated: only 7 of 16 checks could reach a conclusion here; the rest do not apply to this repository or could not read it.
Fit with the Acquirer
not assessableNot rated: structurally not assessable from a repository.
Read more, or read your own
The full report on tinyhumansai/openhuman: every dimension, the questions for the room, and the audit log.
Survey a repository: free on public repositories. Private code is read on your own machine by a collector whose source you can read first.