instar@1.3.1358
⚠ Under reviewCoherence infrastructure for self-evolving AI agents — on the Claude Code or Codex subscription you already have.
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe inspected excerpts show ordinary native-module repair, local commit-signing setup, and intended Telegram API use, without proving malicious behavior. A clean determination would require citation coverage across more distinct files than the response limit permits, so the result remains unresolved.
Decision evidence
public snapshot- package.json runs a postinstall repair script; that script downloads and extracts a platform-specific better-sqlite3 prebuild from its upstream GitHub release.
- GitSync writes a temporary signing probe and invokes ssh-keygen to test local commit signing.
- TelegramAdapter constructs Telegram API requests for api.telegram.org and redacts the bot token in its safe URL.
- PostUpdateMigrator writes a named guard hook under an Instar hooks directory; the inspected excerpt does not establish its activation or broader effects.
- The inspected credential excerpts show OS keychain reads and writes, without establishing a transfer to an unrelated recipient.
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@1.3.1358.
See all versions of instar for other recorded verdicts.
- Review the evidence and your use of instar@1.3.1358 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.