triflux@10.65.0
CLI-first multi-model orchestrator for Claude Code — route tasks to Codex, Antigravity, and Claude
AI Security Review
scanned 11d ago · by lpm-firewall-aiThe automatic setup includes first-party agent configuration and hook registration, creating a concrete lifecycle integration risk. Consent gates and the absence of an inspected exfiltration or foreign control-surface attack support a warning rather than a block.
Decision evidence
public snapshot- package.json runs scripts/setup.mjs automatically after installation.
- The postinstall setup changes Claude settings and ensures Codex hooks are registered; it also cleans up legacy Antigravity hooks.
- The Codex hook setup defines session-start and prompt-submit hooks that run package scripts.
- Setup requires explicit consent for optional installs and Codex hook approval, but the hook registration path also runs during postinstall without that approval.
- The indicated integration locations read or store session, cache, task, or diagnostic data; these excerpts do not establish credential exfiltration or a foreign control-surface attack.
- The setup code identifies Codex hook approval and optional installs as consent-gated, and defers when interaction is unavailable.
- Inspected integration excerpts show local configuration and data paths without a credential send operation or unrelated remote endpoint.
Source & flagged code
25 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgPackage source references shell execution.
Findings
Affected versions and remediation
This report applies to triflux@10.65.0.
See all versions of triflux for other recorded verdicts.
- Review the evidence and your use of triflux@10.65.0 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.