instar@1.3.1421
⚠ Under reviewCoherence infrastructure for self-evolving AI agents — on the Claude Code or Codex subscription you already have.
AI Security Review
scanned 7h ago · by lpm-firewall-aiInspected behavior was consistent with native-module repair, explicit setup, and intended Claude authentication; it did not establish an attack. The complete clean citation checklist exceeds the allowed citation count, so this review cannot certify the package clean.
Decision evidence
public snapshot- package.json runs a postinstall native-module repair script; the inspected script describes repairing better-sqlite3.
- The CLI exposes setup as an explicit command and invokes the setup wizard; inspected setup code writes agent configuration.
- The inspected credential code reads and updates local keychain entries, and OAuth refresh defaults to Claude's token endpoint.
- The inspected better-sqlite3 install hook runs prebuild-install or node-gyp to build its native module.
- The inspected source did not establish a concrete attack or an unrelated credential recipient.
- The full clean review requires more distinct source citations than the output limit permits.
Source & flagged code
29 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@1.3.1421.
See all versions of instar for other recorded verdicts.
- Review the evidence and your use of instar@1.3.1421 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.