review-scrape
Pass
Audited by Gen Agent Trust Hub on Jun 18, 2026
Risk Level: SAFEPROMPT_INJECTION
Full Analysis
- [SAFE]: The skill's primary functionality is to gather and analyze public book review data. All operations, including web scraping and local file management, utilize standard platform tools for their intended research and documentation purposes without evidence of malicious intent.
- [PROMPT_INJECTION]: The skill ingests untrusted data from external sources (Goodreads and Amazon reviews), representing a surface for indirect prompt injection. This is evaluated as safe because the skill treats scraped content as static text for analysis and storage, rather than using it to dynamically construct commands or override agent instructions.
- Ingestion points: Public review text and metadata from Goodreads, Amazon, and various editorial review websites (SKILL.md, Stages 2-4).
- Boundary markers: Scraped text is stored within Markdown code blocks, quotes, and details tags in the final output files (SKILL.md, Stage 6).
- Capability inventory: Access to browser automation tools, shell access via Bash, and file system operations (SKILL.md, allowed-tools).
- Sanitization: No explicit sanitization or filtering of the review text is mentioned, but the skill follows a specific extraction schema for targeted data points.
Audit Metadata