instar@1.3.1331
⚠ Under reviewCoherence infrastructure for self-evolving AI agents — on the Claude Code or Codex subscription you already have.
AI Security Review
scanned 16d ago · by lpm-firewall-aiThe inspected source does not substantiate the scanner's malicious label. Because several flagged agent configuration paths were not traced through callers and consent gates, the remaining risk is unresolved rather than confirmed malicious behavior.
Decision evidence
public snapshot- package.json runs a postinstall script that repairs better-sqlite3; the inspected script uses child processes for a native dependency recovery flow.
- The inspected agent-related excerpts show Claude configuration paths and generated skills or hooks, but do not establish a concrete unconsented install-time control-surface mutation.
- The flagged GitSync source describes project state synchronization and imports guarded filesystem and Git helpers; the inspected excerpt does not show backdoor persistence.
- The supplied flagged paths include additional agent configuration, migration, and GitSync behaviors that were not fully traced to their callers or activation conditions.
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@1.3.1331.
See all versions of instar for other recorded verdicts.
- Review the evidence and your use of instar@1.3.1331 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.