burnboard-cli@0.3.0
Connect local AI coding token usage to BurnBoard
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThis is not an automatic install-time attack, but the explicit command exposes a real unpinned remote-code and persistence capability controlled by a configurable endpoint. The host-data upload compounds that risk.
Decision evidence
public snapshot- The CLI accepts a caller-controlled server URL and invokes a setup script.
- Setup downloads an unpinned agent script from that server, then schedules it to run every five minutes.
- The connect command sends the machine hostname and platform to the selected server.
- The installer removes Claude telemetry environment settings.
- The manifest contains only a prepack script and no install lifecycle hook.
- The risky setup path is reached through explicit CLI commands, not package import.
Source & flagged code
5 flagged · loading sourceSource collects local host identity data and sends it to an external endpoint.
cli.mjsView on unpkg · L4Manifest-reachable source overwrites another installed package with package-defined remote behavior.
Findings
Affected versions and remediation
This report applies to burnboard-cli@0.3.0.
See all versions of burnboard-cli for other recorded verdicts.
- Review the evidence and your use of burnboard-cli@0.3.0 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.