landing-page-optimizer
Pass
Audited by Gen Agent Trust Hub on Sep 15, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONMETADATA_POISONING
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze external landing page URLs and descriptions. \n
- Ingestion points: Prompt examples in SKILL.md explicitly request the analysis of user-provided URLs and descriptions. \n
- Boundary markers: The instructions do not define delimiters or specific safety warnings to prevent the agent from following instructions embedded within the analyzed content. \n
- Capability inventory: The skill defined in SKILL.md contains no tool definitions, script logic, file system access, or network command capabilities. \n
- Sanitization: No sanitization or validation of the ingested external content is mentioned. \n- [METADATA_POISONING]: The skill contains significant metadata inconsistencies. The author is listed as ClawFu in the SKILL.md frontmatter, which contradicts the provided author context of guia-matthieu. Additionally, the Skill Boundaries section incorrectly refers to audio production workflows and audio engineering, which is unrelated to the landing page optimization theme of the skill. These discrepancies represent misleading metadata regarding the skill's scope and authorship.
Audit Metadata