instar@1.3.1382
⚠ Under reviewCoherence infrastructure for self-evolving AI agents — on the Claude Code or Codex subscription you already have.
AI Security Review
scanned 4d ago · by lpm-firewall-aiInspected source excerpts explain the flagged lifecycle and runtime behavior without establishing a concrete attack. The required clean-path coverage exceeds the citation limit, so the package-wide verdict remains unresolved.
Decision evidence
public snapshot- package.json runs a SQLite native-binary repair at postinstall and Husky at prepare.
- The inspected repair script describes testing or rebuilding better-sqlite3 for the active Node version.
- The inspected voice-transcription code selects official Groq or OpenAI endpoints and reads their matching API keys from the environment.
- The inspected GitSync code writes public signing keys to Git’s allowed-signers file.
- Post-update code installs a generated credential prompt guard into Instar’s hooks directory.
- The review’s required clean-path citation set exceeds the 256-citation output limit, so this inspection cannot establish package-wide clean coverage.
Source & flagged code
17 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgPackage source references child process execution.
dist/threadline/PipeSessionSpawner.jsView on unpkg · L15Package source references shell execution.
Findings
Affected versions and remediation
This report applies to instar@1.3.1382.
See all versions of instar for other recorded verdicts.
- Review the evidence and your use of instar@1.3.1382 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.