astrofy-structured-data
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONEXTERNAL_DOWNLOADS
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes visible website content to generate structured data, which represents an indirect prompt injection surface as the source content could be untrusted. However, the skill explicitly mandates sanitization by escaping HTML script tags and validating factual correspondence.\n
- Ingestion points:
SKILL.mdandreferences/technical.md(reading visible content and JSON-LD).\n - Boundary markers: The skill instructs the agent to "Escape conteúdo inserido em script HTML" and "Valide a sintaxe e compare cada afirmação com o texto visível".\n
- Capability inventory: Command execution via the
astrofy checktool.\n - Sanitization: Mandatory use of
JSON.stringifyand specific character escaping (transforming<to\\u003c) to prevent XSS.\n- [COMMAND_EXECUTION]: The skill utilizes a CLI tool namedastrofywith the commandastrofy check --category structuredto perform validation of the generated JSON-LD files.\n- [EXTERNAL_DOWNLOADS]: The skill references documentation and guidelines from Schema.org and Google Developers. These are well-known technology organizations and the references are used neutrally for configuration and policy adherence.
Audit Metadata