@clawos-dev/clawd@0.2.683
Standalone clawd daemon — Claude Code (and future Codex) session server over WebSocket
AI Security Review
scanned 4d ago · by lpm-firewall-aiThe source supports a warning for default-on operational log egress, including identifiers and potentially sensitive diagnostic text. The scanner’s similarity hints and Claude configuration matches did not establish malicious behavior.
Decision evidence
public snapshot- The manifest exposes a CLI; its declared scripts are build, development, type-checking, and test tasks, with no install lifecycle hook.
- The flagged Claude paths show product behavior: the CLI appends turn-duration metadata only to an existing session log, and the plugin references a project collection script in its publish flow.
- Daemon log shipping defaults to on and selects the Clawd Alibaba log endpoint unless disabled or configured off.
- The log client packages records and sends them by POST to its configured endpoint.
- Log records include owner and device identifiers and message text; the CLI also logs up to 500 characters of tool stderr, creating a risk that sensitive diagnostic content is included.
Source & flagged code
14 flagged · loading sourcePackage source references child process execution.
dist/app-builder-plugin/mcp-server.cjsView on unpkg · L30858Source file is highly similar to a previously finalized malicious package; route for source-aware review.
dist/app-builder-plugin/mcp-server.cjsView on unpkgPackage source references a known benign dynamic code generation pattern.
dist/app-builder-plugin/mcp-server.cjsFindings
Affected versions and remediation
This report applies to @clawos-dev/clawd@0.2.683.
See all versions of @clawos-dev/clawd for other recorded verdicts.
- Review the evidence and your use of @clawos-dev/clawd@0.2.683 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.