page-cro
Pass
Audited by Gen Agent Trust Hub on Jul 10, 2026
Risk Level: SAFECOMMAND_EXECUTIONEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill utilizes a local Python script
scripts/conversion_audit.pyto automate the analysis of HTML documents. This script is intended to be invoked by the agent to generate conversion scores and reports. - [EXTERNAL_DOWNLOADS]: The included script
scripts/conversion_audit.pyuses theurllib.requestlibrary to download HTML content from external websites when a URL is provided. This is a core functional feature for analyzing live pages. - [PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes data from external sources that may be controlled by third parties.
- Ingestion points: Data enters the agent's context through the output of
scripts/conversion_audit.py, which reads from local HTML files or remote URLs. - Boundary markers: There are no explicit markers or instructions provided to the agent to disregard instructions found within the analyzed HTML content.
- Capability inventory: The agent has the ability to execute scripts, read local files, and make network requests.
- Sanitization: While the script parses the HTML structure safely, it does not sanitize the extracted text for potential instructions directed at the AI agent.
Audit Metadata