@ngockhoale/ukit@2.7.8
Install/update an index-first AI workspace for Claude Code, OpenAI Codex and omp (Oh My Pi).
AI Security Review
scanned 15d ago · by lpm-firewall-aiThe package exposes an explicit CLI without an install hook, while its shipped agent hook can influence Stop handling. This supports a warning for package-owned agent capability, not a confirmed attack.
Decision evidence
public snapshot- The manifest exposes UKit through an explicit CLI and has no install lifecycle hook.
- The bundled task watchdog delegates a Stop event to a coordinator that can affect continuation decisions.
- The watchdog starts the coordinator as a child process when the coordinator file exists.
- The watchdog is part of a package-owned agent workspace template, and its Stop behavior is activated through the agent hook system.
- The inspected sources show project settings preservation and project-scoped indexing behavior; no credential collection or external endpoint was identified.
Source & flagged code
7 flagged · loading sourcePackage source references dynamic require/import behavior.
scripts/bench/runGold.mjsView on unpkg · L110Package source references weak cryptographic algorithms.
src/core/memory/migrate.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @ngockhoale/ukit@2.7.8.
See all versions of @ngockhoale/ukit for other recorded verdicts.
- Review the evidence and your use of @ngockhoale/ukit@2.7.8 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.