Read 2026-11-03
nanocoai/nanoclaw: technical due diligence, read from its repository
What Borehole read in nanocoai/nanoclaw's code and history at 8797ad3: 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 | strong |
| Security & Reliability Posture | adequate |
| Observability & Operations | strong |
| Fit with the Acquirer | not assessable |
What the repository says
CONTRIBUTING.md:201: states a policy on AI-generated contributions
Architecture & Codebase — including AI provenance
strongsetup/auto.ts is 1,806 lines, 15× this codebase's median file
adequateThe median source file here is 117 lines. 5 files exceed 1,170 lines, the largest being 1,806. 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.
setup/auto.ts:1 — 1,806 lines (15× median) |
src/container-runner.ts:1 — 1,672 lines (14× median) |
container/agent-runner/src/poll-loop.ts:1 — 1,386 lines (11× median) |
src/channels/chat-sdk-bridge.ts:1 — 1,305 lines (11× median) |
setup/registry-login.ts:1 — 1,266 lines (10× median) |
Engineering Process & SDLC Maturity
strongNo finding.
Product & Engineering Maturity
strongNo finding.
Security & Reliability Posture
adequate16 credential-shaped strings 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. 5 other matchs set aside: expired tokens, licence tokens a vendor verifies with a public key, or placeholder text where a key would go. None grants access to anything.
| GitHub token, where fakes are usual |
| 4 cited: Slack token, where fakes are usual |
The build runs third-party code it does not pin
adequate8 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.
.claude/skills/add-iron-proxy/scripts/install-claude.sh:10 — download piped into a shell |
18 shell commands built from strings
adequate18 places run a command through a shell (`shell=True`, `os.system`, `exec` with a template) rather than passing arguments. Where any part comes from outside, it is command injection. Ask where each one's input comes from.
setup/platform.ts:60 — command run through a shell |
setup/platform.ts:66 — command run through a shell |
setup/platform.ts:70 — command run through a shell |
3 locked dependencies have published vulnerabilities
adequate3 of 310 locked dependencies match 4 published advisories in OSV. 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.
pnpm-lock.yaml — @vitest/mocker 4.1.11: GHSA-qxg6-89mg-rqv8 (moderate), fixed in 4.1.12; a development dependency |
pnpm-lock.yaml — minimatch 10.2.5: GHSA-g3hv-75wg-m249 (moderate), fixed in 11.0.1; a development dependency |
pnpm-lock.yaml — minimatch 3.1.5: GHSA-g3hv-75wg-m249 (moderate), fixed in 11.0.1; a development dependency |
18 advisories against this project in the last 24 months, 7 high or critical, 8 fixed in no release yet
adequateThe newest was published on 2026-10-13. 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 3 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 2.4.0. Answer lives in: the release you run, set against each fixed version. GHSA-628g-jj3x-6mp9 (high), GHSA-8w49-2gv4-3x6w (high), GHSA-vwj5-j44r-v364 (high) and 5 more affect 2.4.0, the newest release, and not the surveyed commit: their fixes are in the surveyed commit and in no release. Answer lives in: the version you run. The record is incomplete: NVD: 1 record names nanoclaw with nothing tying it to this repository (CVE-2026-7875).
| 2026-10-13: GHSA-628g-jj3x-6mp9 (high), fixed in the surveyed commit and in no release: OneCLI setup downloaded its installer over plain HTTP and ran it |
| 2026-10-13: GHSA-8w49-2gv4-3x6w (high), fixed in the surveyed commit and in no release: Mattermost setup used a server-supplied user ID in a setup command without checking it |
| 2026-10-13: GHSA-vwj5-j44r-v364 (high), fixed in the surveyed commit and in no release: Session inbox directory was not checked again between the attachments of one message |
| 2026-10-13: GHSA-g3gp-6777-8p9g (moderate), fixed in the surveyed commit and in no release: Non-admin group members could run /clear and /upload-trace by adding a suffix |
| 2026-10-13: GHSA-9j3v-9hwg-v896 (moderate), fixed in the surveyed commit and in no release: Task run-log read built a host file path from a series id without checking it |
Observability & Operations
strongNo error reporter is wired in
adequateNothing in the service initialises an error reporter (Sentry, Rollbar, Bugsnag, Honeybadger or an APM agent). An exception in production reaches a person only if someone is reading the logs, which usually means a customer reports it first.
| no error reporter initialised |
Fit with the Acquirer
not assessableNot rated: structurally not assessable from a repository.
Read more, or read your own
The full report on nanocoai/nanoclaw: 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.