@ngockhoale/ukit@2.3.13
Install/update an index-first AI workspace for Claude Code, OpenAI Codex, OpenCode, and omp (Oh My Pi).
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is not concrete npm lifecycle malware, but it installs an AI-agent control surface that automatically broadens local command permissions. The user-invoked trigger reduces maliciousness confidence but does not remove the capability risk.
Static reason
No blocking static signals were detected.
Trigger
The user runs ukit install and then Claude invokes a Bash tool call.
Impact
Later agent Bash commands may receive broader local permission without a separate approval prompt.
Mechanism
Project-level AI-agent permission mutation through an event hook.
AI rationale
This is not concrete npm lifecycle malware, but it installs an AI-agent control surface that automatically broadens local command permissions. The user-invoked trigger reduces maliciousness confidence but does not remove the capability risk.
Evidence
src/cli/commands/install.jstemplates/.claude/settings.jsontemplates/.claude/hooks/auto-allow-bash.sh.claude/settings.json.claude/settings.local.json.claude/hooks/auto-allow-bash.sh.claude/ukit/permission-usage.json.claude/ukit/permission-audit.log
Decision evidence
public snapshotAI called this Suspicious at 90.0% confidence as Dangerous Capability with low false-positive risk.
Evidence for warning
- The explicit install command writes a managed AI-workspace configuration into the current project.
- The Claude template sets bypass-permissions mode and registers hooks for agent tool events.
- The Bash hook automatically adds allow rules for encountered command binaries.
- The flagged sensitive-data hook blocks detected secrets, while the vision hook only invokes a local extractor.
Evidence against
- package.json has no preinstall, install, or postinstall script.
- No source-inspected network request or credential exfiltration was found.
- Configuration mutation occurs through the user-invoked ukit install command, not automatically on npm installation.
Behavioral surface
ChildProcessCryptoEnvironmentVarsFilesystemShell
HighEntropyStrings
Source & flagged code
7 flagged · loading sourcetemplates/.claude/hooks/sensitive-data-guard.shView file
•path = templates/.claude/hooks/sensitive-data-guard.sh
kind = payload_in_excluded_dir
sizeBytes = 12468
magicHex = [redacted]
High
Payload In Excluded Dir
Package hides binary, compressed, or executable-looking payloads in test/fixture/hidden paths.
templates/.claude/hooks/sensitive-data-guard.shView on unpkg•path = templates/.claude/hooks/sensitive-data-guard.sh
kind = build_helper
sizeBytes = 12468
magicHex = [redacted]
Medium
Ships Build Helper
Package ships non-JavaScript build or shell helper files.
templates/.claude/hooks/sensitive-data-guard.shView on unpkgpackage.jsonView file
•stage = ast_semantic_analysis; reason = ast_parse_error; limitedFiles = 1
High
Findings
2 High7 Medium3 Low
HighPayload In Excluded Dirtemplates/.claude/hooks/sensitive-data-guard.sh
HighSemantic Analysis Limitedpackage.json
MediumEnvironment Vars
MediumShips Build Helpertemplates/.claude/hooks/sensitive-data-guard.sh
MediumStructural Risk Force Deep Review
MediumAi Review Evidencesrc/cli/commands/install.js
MediumAi Review Evidencetemplates/.claude/settings.json
MediumAi Review Evidencetemplates/.claude/settings.json
MediumAi Review Evidencetemplates/.claude/hooks/auto-allow-bash.sh
LowScripts Present
LowFilesystem
LowHigh Entropy Strings
Affected versions and remediation
This report applies to @ngockhoale/ukit@2.3.13.
See all versions of @ngockhoale/ukit for other recorded verdicts.
- Review the evidence and your use of @ngockhoale/ukit@2.3.13 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.