image-optimization
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFE
Full Analysis
- [SAFE]: The skill provides legitimate image optimization guidance and auditing capabilities without any detected security risks.
- [INDIRECT_PROMPT_INJECTION]: The skill possesses an inherent attack surface because it processes untrusted HTML data from user-provided files or URLs. The audit script (
scripts/audit-images.mjs) parses this content to identify image optimization issues. While this data could contain malicious instructions meant to influence the agent, the risk is mitigated by the skill's specific focus and the standard nature of web auditing tools. - Ingestion points:
scripts/audit-images.mjsreads user-provided HTML files or fetches remote URLs. - Boundary markers: None explicitly defined for untrusted data separation.
- Capability inventory: The script performs network HEAD requests and the agent is capable of file system operations for patching.
- Sanitization: The script uses regular expressions for extraction but does not execute the extracted content.
Audit Metadata