open-walnut@0.6.8
Personal AI powered by Claude — task management, Claude Code sessions, memory system, and web UI
AI Security Review
scanned 12d ago · by lpm-firewall-aiThe install hook makes a concrete, version-checked change to an AI agent safety prompt by removing its malware-related safeguard. The source supports a warning for this capability change, but not a block for broad control-surface hijacking or other concrete attack behavior.
Decision evidence
public snapshot- package.json runs scripts/postinstall.mjs during installation.
- The postinstall hook changes the installed Claude Agent SDK cli.js after checking the patch context.
- The shipped patch removes a system reminder telling the SDK to consider malware risks and refuse to improve malware.
- The flagged agent-related excerpts are configuration paths or context-file lookups, not evidence of broad agent-control writes.
- The flagged web assets contain syntax-highlighting code, and the flagged daemon service excerpt contains WebSocket protocol code.
- The SDK edit is narrowly scoped to a version-matched dependency file and skips when the expected patch context is absent.
- The inspected excerpts do not establish credential collection, exfiltration, or remote payload execution.
Source & flagged code
17 flagged · loading sourcePackage defines install-time lifecycle scripts.
package.jsonView on unpkgInstall-time lifecycle script is not statically allowlisted and needs review.
package.jsonView on unpkgPackage source references child process execution.
dist/plugin-store/walnut-rhythm/dist/server.mjsView on unpkg · L8Findings
Affected versions and remediation
This report applies to open-walnut@0.6.8.
See all versions of open-walnut for other recorded verdicts.
- Review the evidence and your use of open-walnut@0.6.8 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.