eve@0.72.0
Filesystem-first framework for durable backend AI agents that run anywhere.
AI Security Review
scanned 16d ago · by lpm-firewall-aiThe inspected findings are consistent with the package’s agent framework and do not establish credential exfiltration or other concrete attack behavior. The CLI behavior is user-invoked, and the inspected Vercel OAuth endpoints are restricted to Vercel.
Decision evidence
public snapshot- package.json identifies a public package entrypoint and CLI, consistent with an agent framework.
- The package scripts shown contain build and development tasks, not install lifecycle hooks.
- The CLI loads its application entrypoint when invoked; this supports user-triggered execution.
- Vercel OAuth endpoints are restricted to vercel.com and api.vercel.com.
- The flagged HTTP Basic file compares supplied credentials with configured credentials using timing-safe comparison.
- The flagged mock fixture file builds mock tool inputs and fixture replies; the flagged OpenAI file imports provider SDK code.
Source & flagged code
10 flagged · loading sourceSource spawns a local helper that also contains network and dynamic execution context; review data flow before blocking.
bin/eve.jsView on unpkg · L2Package source references dynamic require/import behavior.
bin/eve.jsFindings
Affected versions and remediation
This report applies to eve@0.72.0.
See all versions of eve for other recorded verdicts.
- Review the evidence and your use of eve@0.72.0 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.