@ngockhoale/ukit@3.4.52
Install/update an index-first AI workspace for Claude Code, OpenAI Codex and omp (Oh My Pi).
AI Security Review
scanned 1d ago · by lpm-firewall-aiThe source supports conditional forwarding of an existing Claude credential to a separately configurable decision endpoint without recipient binding. This warrants a warning, but the runtime feature must be enabled and the source does not establish malicious use.
Decision evidence
public snapshot- When the decision plane is enabled, task routing can call the decision service; it defaults to off when its stage is absent or malformed.
- If no decision-specific key is configured, the client falls back to credentials resolved from ANTHROPIC_* environment variables or Claude settings.
- The decision client sends its bearer or API key to the resolved decision endpoint, which may be configured independently of the credential’s intended service.
- The manifest declares no npm lifecycle hooks; the decision-plane request requires runtime activation.
- The package describes the default endpoint as its UNIC decision provider, and sensitive state values are filtered before requests.
Source & flagged code
12 flagged · loading sourcePackage source references dynamic require/import behavior.
src/core/runtimeConfig.jsView on unpkg · L9Source file is highly similar to a previously finalized malicious package; route for source-aware review.
scripts/release/verify-release.mjsView on unpkgFindings
Affected versions and remediation
This report applies to @ngockhoale/ukit@3.4.52.
See all versions of @ngockhoale/ukit for other recorded verdicts.
- Review the evidence and your use of @ngockhoale/ukit@3.4.52 before allowing it. Restrict the permissions described in this report.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.