@proagentstore/cli@0.4.201
CLI for creating, publishing, and running ProAgentStore agents
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe CLI automatically uploads bounded runtime stderr output in crash reports to its own service without a clear consent step. Because the report may contain tool output, this is suspicious undisclosed collection rather than a confirmed credential theft or unrelated-host exfiltration.
Decision evidence
public snapshot- When the local runtime exits unexpectedly, the CLI captures its stderr tail for a crash report and attempts delivery before exiting.
- The CLI POSTs queued crash reports to ProAgentStore's `/v1/errors/runner-crashes` endpoint using the user's session token.
- Crash reports can include the last 20 output lines, with each line capped at 300 characters; those lines may contain user or tool output.
- The manifest declares no install, preinstall, or postinstall hooks; the package is an explicitly invoked CLI.
- The package's core purpose includes running a local browser runtime and connecting it to ProAgentStore, so its ordinary network and process use is expected.
Source & flagged code
10 flaggedPackage source references child process execution.
dist/browser-runner/site-builder/runtime.jsView on unpkg · L32Package source references shell execution.
dist/browser-runner/coding/windows-runner-live.jsView on unpkg · L59Findings
Affected versions and remediation
This report applies to @proagentstore/cli@0.4.201.
See all versions of @proagentstore/cli for other recorded verdicts.
- Review the evidence and your use of @proagentstore/cli@0.4.201 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.