@agentvault/agentvault@0.23.57
End-to-end encrypted messaging for AI agents: connects an agent to its owner, rooms and other agents on AgentVault.
AI Security Review
scanned 5d ago · by lpm-firewall-aiThe package’s observability feature automatically collects full URLs from outbound requests and uploads them with the AgentVault device credential. This is runtime telemetry with broader collection than the README discloses, so it warrants a warning.
Decision evidence
public snapshot- The OpenClaw entry point installs a fetch interceptor and forwards observed HTTP request URLs to channel telemetry.
- Telemetry spans retain the complete request URL as `url.full`.
- After connecting, the channel creates a telemetry reporter using its device JWT; it posts buffered spans to `/api/v1/telemetry/ingest`.
- The README mentions telemetry for message operations but does not disclose collection of URLs from outbound HTTP requests.
- The package documents telemetry as feeding its trust scoring engine and observability dashboard.
- The manifest has no install, preinstall, or postinstall hook; `prepublishOnly` only runs the build.
Source & flagged code
10 flaggedA single source file combines environment access, network access, and code or shell execution; review context before blocking.
dist/cli.jsView on unpkg · L93622Findings
Affected versions and remediation
This report applies to @agentvault/agentvault@0.23.57.
See all versions of @agentvault/agentvault for other recorded verdicts.
- Review the evidence and your use of @agentvault/agentvault@0.23.57 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.