copywriting
Pass
Audited by Gen Agent Trust Hub on Nov 12, 2026
Risk Level: SAFECOMMAND_EXECUTIONINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [COMMAND_EXECUTION]: The skill provides a set of shell commands and inline Python scripts used for performing 'readiness checks' on landing pages. These commands utilize
curlto fetch external website content and local Python logic to verify that the copy meets brand and platform requirements. - [INDIRECT_PROMPT_INJECTION]: During the auditing of landing pages, the skill ingests untrusted HTML content from external URLs. This creates a surface for indirect prompt injection, as the agent processes data from third-party sites. The risk is minimized as the provided tools focus on structural and lexical analysis rather than interpreting the content as instructions.
- [EXTERNAL_DOWNLOADS]: The documentation references and suggests the use of several external packages and companion skills, including
sshlg-skills,sheleg-design-skill, andtask-pipeline. These resources are identified as vendor-owned (ssheleg/sshlg) and are used for tool discovery and extending system capabilities. - [DYNAMIC_EXECUTION]: The skill's auditing workflow involves executing transient Python scripts defined within the reference documentation. These scripts are run via shell heredocs and are used for parsing HTML content, relying exclusively on the Python standard library.
Audit Metadata