superlocalmemory@4.1.15
Local-first agent memory with MCP and an agent-native CLI. Documented clients include Claude Code, Cursor, and Windsurf.
AI Security Review
scanned 1mo ago · by lpm-firewall-aiThe package contains a remote runtime-loader and a first-party agent plugin with automatic hooks, so its effective behavior cannot be fully verified from this npm snapshot. Its lifecycle code does not establish the concrete malicious chain required for a publish block.
Decision evidence
public snapshot- The npm postinstall hook runs automatically.
- Postinstall creates a private virtual environment and uses pip to install a separately fetched Python package of the same name and version.
- The included first-party Claude plugin starts commands on session start and after prompts and tool use.
- The plugin bootstrap installs the pinned Python package into persistent plugin storage.
- The npm lifecycle code only targets its package-owned virtual environment and does not write host agent configuration.
- No credential harvesting or external exfiltration endpoint is present in the inspected shipped source.
- IDE configuration files are templates; postinstall does not install them.
Source & flagged code
10 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 unpkgSource file is highly similar to a previously finalized malicious package; route for source-aware review.
Findings
Affected versions and remediation
This report applies to superlocalmemory@4.1.15.
See all versions of superlocalmemory for other recorded verdicts.
- Review the evidence and your use of superlocalmemory@4.1.15 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.