instar@1.3.1376
Coherence 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-aiThe source shows an update migration that installs project-scoped Codex hooks and automatically accepts the trust prompt by default. This supports a warning about first-party agent extension lifecycle risk, but does not establish malicious behavior or credential exfiltration.
Decision evidence
public snapshot- package.json declares a postinstall helper and a prepare hook; the inspected migration is invoked later by server startup.
- Server startup passes the project and state directories to a post-update migrator.
- On update, the migrator registers hooks when Codex is enabled and auto-arms them by default when Codex is found; setting autoArmHooks to false disables this step.
- The hook installer writes project-level Codex settings, replaces Instar-owned hooks and preserves user-added hooks. The arm flow accepts the changed-hook trust prompt automatically.
- The other flagged excerpts concern project skills, framework detection, local keychain access, and a quota notification callback; they do not show an unrelated credential transfer.
- The hook behavior runs during an update migration, not from the package's postinstall hook.
- The installed hooks are scoped to the project and existing user-added hooks are preserved.
- The inspected credential code reads and writes macOS Keychain data; the cited excerpts do not show sending it to an unrelated recipient.
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@1.3.1376.
See all versions of instar for other recorded verdicts.
- Review the evidence and your use of instar@1.3.1376 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.