@punks/cli@3.2.1
Harness Intelligence scaffolding CLI
AI Security Review
scanned 3mo ago · by lpm-firewall-aiNot malicious: no npm install hook or concrete malicious chain was found. Warn because the explicit scaffold workflow mutates AI-agent control surfaces and registers command hooks.
Decision evidence
public snapshot- Explicit scaffold projection configures project-local Codex command hooks.
- Claude projection mirrors hooks and merges project settings.
- Bundled scaffold can modify AI-agent control surfaces after a user runs its CLI.
- package.json exposes only bin entrypoints; no npm lifecycle hook is declared.
- Hook targets are project-local and named as formatting/scaffold-update checks.
- No credential harvesting, payload download, or exfiltration was confirmed.
Source & flagged code
12 flaggedPackage source references child process execution.
dist/baseline/scaffold-baseline/skills/misc/timesheet/scripts/git-timesheet-evidence.mjsView on unpkg · L1Package source references dynamic require/import behavior.
dist/baseline/scaffold-baseline/skills/frameworks/expo/expo-ui/scripts/list-components.jsView on unpkg · L15Findings
Affected versions and remediation
This report applies to @punks/cli@3.2.1.
See all versions of @punks/cli for other recorded verdicts.
- Review the evidence and your use of @punks/cli@3.2.1 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.