review-intelligence-digest
Pass
Audited by Gen Agent Trust Hub on Jul 20, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill has an attack surface for indirect prompt injection because it ingests and analyzes untrusted data from public review websites.
- Ingestion points: External review content from G2, Capterra, and Trustpilot is ingested into the agent context via the
scrape_reviews.pyscript. - Boundary markers: The skill instructions do not define explicit delimiters or use specific 'ignore embedded instructions' prompts to isolate the scraped content from the agent's operational logic.
- Capability inventory: The agent has the capability to execute local Python scripts and write analysis reports to the file system in the
clients/directory. - Sanitization: No sanitization or filtering of the scraped review text is mentioned before the data is passed to the LLM for categorization and clustering.
- [COMMAND_EXECUTION]: The skill executes a local Python script
skills/review-scraper/scripts/scrape_reviews.pyto perform data collection. This is an intended part of the skill's functionality and utilizes a local path, which is consistent with the platform's modular skill architecture.
Audit Metadata