claude-recall@0.40.0
Persistent memory for Claude Code and Pi with native Skills integration, automatic capture, failure learning, and project scoping
AI Security Review
scanned 23d ago · by lpm-firewall-aiAutomatic installation-time changes to existing agent hooks warrant a warning under the guarded extension lifecycle policy. Package-specific filtering and backups support withholding a malicious verdict and publication block.
Decision evidence
public snapshot- The npm postinstall script automatically invokes repair with automatic approval and all scopes.
- The CLI passes automatic approval and scope directly to the repair implementation.
- Repair scans global Claude settings and project settings under the user's home.
- Repair rewrites recognized broken or incompatible Claude Recall hook commands and writes settings with backups.
- Repair filters for Claude Recall commands and leaves unrelated hooks alone.
- Changes require an existing recognized hook, a resolvable package binary, and a parsed hook identifier.
- The inspected setup installer requires an explicit CLI option; no confirmed exfiltration or remote payload execution was established.
Source & flagged code
20 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 unpkgFindings
Affected versions and remediation
This report applies to claude-recall@0.40.0.
See all versions of claude-recall for other recorded verdicts.
- Review the evidence and your use of claude-recall@0.40.0 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.