open-walnut@0.6.15
Personal AI powered by Claude — task management, Claude Code sessions, memory system, and web UI
AI Security Review
scanned 2d ago · by lpm-firewall-aiThe package automatically modifies a foreign AI agent SDK during installation and suppresses its file-read safety reminder. This is an unconsented change to the agent's control behavior, supporting a block.
Decision evidence
public snapshot- package.json runs a postinstall hook and pins the Claude Agent SDK to version 0.2.50.
- The postinstall hook calls the patch routine automatically, without a user consent prompt.
- The routine rewrites the installed SDK CLI, and its message says the reminder is disabled.
- The patch changes file-read results from conditionally appending a reminder to appending nothing.
- The patch contains a file-read reminder directing the SDK to refuse improving malware.
- The patch is applied only when the installed SDK version matches its target.
Source & flagged code
26 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@0.6.15.
See all versions of open-walnut for other recorded verdicts.
- Avoid installing open-walnut@0.6.15. Remove it from direct dependencies and check your lockfile for transitive copies.
- Choose an independently verified alternative or release. This report does not establish that other versions are safe.
- If this version ran, investigate the affected machine and build environment. Rotate credentials it could access and rebuild from a trusted environment.
Evidence last updated: .
Think this verdict is wrong? Dispute it on GitHub or read how reports are produced.