@andy-toolforge/knowledge-base
Filesystem-first knowledge base facade for AI agents: JSON store with CRUD/search, optional best-effort Supermemory/Serena adapters, and a management skill installed into client projects.
This report shows the evidence for @andy-toolforge/knowledge-base@0.1.1, the only known version.
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is a concrete unconsented postinstall mutation of a foreign AI-agent control surface. The absence of exfiltration does not remove that install-time control-surface risk. Product guard normalized a concrete AI-agent control hijack publish_block to the blockable dangerous-capability shape.
Decision evidence
public snapshot- npm postinstall executes skills/postinstall.js.
- The lifecycle script delegates skill installation to @andy-toolforge/core.
- README documents copying a management skill into the client project's .opencode/skills control surface.
- No network, credential harvesting, or payload download code is present.
- Runtime storage is a local JSON file; external CLIs run only when kb_add is invoked.
Source & flagged code
4 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 unpkgThe lifecycle script delegates skill installation to @andy-toolforge/core.
skills/postinstall.jsView on unpkg · L1Findings
Affected versions and remediation
This report applies to @andy-toolforge/knowledge-base@0.1.1.
The versions list below shows the verdict for each recorded version.
- Avoid installing @andy-toolforge/knowledge-base@0.1.1. 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.