The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The scripts/conversion_audit.py script performs network read operations using urllib.request.urlopen to fetch HTML content from user-provided URLs. This behavior is expected and necessary for the skill's purpose of auditing marketing pages. \n- [COMMAND_EXECUTION]: The skill executes a local Python script, scripts/conversion_audit.py, which is included in the skill package. The script performs HTML parsing and regex-based analysis to score page elements. \n- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection vulnerabilities. Ingestion points: Untrusted data enters the agent context through the HTML content fetched from external URLs or read from local files in scripts/conversion_audit.py. Boundary markers: The skill does not employ explicit delimiters or instructions to ignore embedded commands within the analyzed HTML. Capability inventory: The skill involves local script execution and network read capabilities. Sanitization: There is no evidence of content sanitization or instruction filtering for the data ingested during the audit process, which is then used to generate agent responses.

page-cro