instar@1.3.1349
Coherence infrastructure for self-evolving AI agents — on the Claude Code or Codex subscription you already have.
AI Security Review
scanned 12d ago · by lpm-firewall-aiThe version-gated migration installs and registers a hook that automatically allows PermissionRequest events, creating a real agent permission risk. This supports a warning for first-party agent extension setup, but not a malicious block.
Decision evidence
public snapshot- The package’s postinstall runs a SQLite repair script; the agent permission change occurs in a separate server-start migration.
- On startup, a version change triggers the post-update migration.
- The migration adds a PermissionRequest handler to the Claude settings hooks.
- That handler returns an allow decision for all PermissionRequest events.
- The migration writes the updated settings file, and the package-owned hook is installed under the project’s Instar hooks directory.
- The permission change is first-party setup performed by a version-gated migration, not an npm install hook that mutates a foreign agent configuration.
- The package describes the hook as a way to unblock subagent tool use; the source does not show credential collection, exfiltration, or remote payload execution.
Source & flagged code
18 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@1.3.1349.
See all versions of instar for other recorded verdicts.
- Review the evidence and your use of instar@1.3.1349 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.