@inizioevoke/astro-core@3.2.1
AI Security Review
scanned 2mo ago · by lpm-firewall-aiThis is an unconsented install-time mutation of an AI-agent control surface, but it is limited to the package's own named skill and contains no confirmed payload delivery or exfiltration. It warrants a warning under the first-party agent-extension lifecycle policy.
Decision evidence
public snapshot- The package automatically runs a postinstall script.
- The script deletes and replaces a skill directory in the consuming project's .claude/skills control surface.
- The installed skill instructs Claude to activate for broad Astro component requests, not only explicit package requests.
- The lifecycle script has no network, shell, credential, or environment-variable access.
- The manifest has no runtime dependency on another release of this package.
- The bundled PDF viewer is imported as a local component asset; no install-time PDF execution occurs.
Source & flagged code
5 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 dynamic require/import behavior.
Findings
Affected versions and remediation
This report applies to @inizioevoke/astro-core@3.2.1.
See all versions of @inizioevoke/astro-core for other recorded verdicts.
- Review the evidence and your use of @inizioevoke/astro-core@3.2.1 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.