halfcycle@0.3.24
Your coding agents are the engineers. Halfcycle is the rest of the team: product owner, architect, planner and independent reviewer for Claude Code.
AI Security Review
scanned 28d ago · by lpm-firewall-aiThe package has a real, package-owned agent setup capability that writes Claude hooks and settings when its CLI flow is run, so a warning is appropriate. The manifest points to CLI and library entrypoints, and inspected evidence supports explicit setup rather than npm lifecycle mutation or a confirmed attack.
Decision evidence
public snapshot- The package exposes a CLI and library entrypoint, with no npm install lifecycle hook in its manifest.
- The CLI has an allowlist for writes to Claude hooks, settings, and Halfcycle project files.
- The CLI calls the installer after creating or joining an engagement.
- The credential declaration describes a Claude hook that loads engagement credentials into the runner environment.
- The library uses the same bounded write allowlist.
- The bundled executable invokes Git with specified arguments and a working directory.
- The agent configuration writes are part of the package’s explicit CLI setup flow; no unconsented npm lifecycle mutation was identified.
- No source evidence inspected established credential theft, unrelated data exfiltration, or remote code execution.
Source & flagged code
6 flaggedPackage source references child process execution.
bin/bin.bundle.mjsView on unpkg · L15328Findings
Affected versions and remediation
This report applies to halfcycle@0.3.24.
See all versions of halfcycle for other recorded verdicts.
- Review the evidence and your use of halfcycle@0.3.24 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.