triflux@10.57.1
CLI-first multi-model orchestrator for Claude Code — route tasks to Codex, Antigravity, and Claude
AI Security Review
scanned 13d ago · by lpm-firewall-aiSource inspection confirms automatic postinstall setup that changes user-level agent configuration and synchronizes package-owned extensions without an interactive prompt. This warrants a lifecycle-risk warning, while the inspected code does not establish a concrete malicious attack.
Decision evidence
public snapshot- package.json runs scripts/setup.mjs during postinstall, and the installed package invokes runDeferred automatically.
- Postinstall cannot prompt, but setup calls syncSkills before the consent steps.
- syncSkills copies package skill files into the user-level Claude skills directory; setup also invokes Codex profile and hook setup.
- This is automatic setup of package-owned agent extensions, a guarded lifecycle risk.
- The source routes plugin installation and Codex hook approval through separate consent steps.
- No credential collection or unrelated network recipient was identified in the inspected setup path.
Source & flagged code
25 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
bin/tfx-live.mjsView on unpkgFindings
Affected versions and remediation
This report applies to triflux@10.57.1.
See all versions of triflux for other recorded verdicts.
- Review the evidence and your use of triflux@10.57.1 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.