superlocalmemory@4.2.11
The governed memory brain for AI agents: one local memory for Claude Code, Codex, Cursor, Copilot, ChatGPT and any MCP client. Honest recall, documents and pictures, roles, audit and GDPR erasure.
AI Security Review
scanned 17h ago · by lpm-firewall-aiThe flagged install and runtime behaviors align with the package’s local-memory and document-generation functions, with optional setup and destructive commands gated by user action. The scanner’s environment match is part of docx’s generated browser shim; no concrete attack behavior was found in the inspected source.
Decision evidence
public snapshot- The manifest’s install hook creates a package-local Python runtime and installs the matching SuperLocalMemory release; the CLI wrapper uses that runtime, and the uninstall hook only inspects data size.
- Optional media setup requires an explicit flag or environment setting, or a default-no terminal prompt; the installer records a request rather than downloading media.
- The flagged skill files route user-invoked commands to the memory CLI. Profile switching and reset operations ask for confirmation.
- The declared docx dependency exposes document-generation bundles. Its flagged environment assignment appears in a generated browser-process shim.
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@4.2.11.
See all versions of superlocalmemory for other recorded verdicts.
- Review the evidence and your use of superlocalmemory@4.2.11 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.