autonomous-seo-agent
Pass
Audited by Gen Agent Trust Hub on Oct 12, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill scans untrusted HTML files within a repository to detect SEO issues and generate patches. This creates an attack surface where malicious instructions embedded in the HTML could influence the agent's logic for generating pull request descriptions or applying 'draft' fixes (where the agent derives copy from page text).
- Ingestion points: HTML files located in the directory specified by the
--dirargument. - Boundary markers: No specific boundary markers or instructions to ignore embedded content are defined in the skill instructions.
- Capability inventory: The skill can rewrite local files (
--apply), create git branches, commit changes, and open pull requests (--pr). - Sanitization: No explicit sanitization or validation of the processed HTML content is mentioned before it is used for decision-making or PR body generation.
- [COMMAND_EXECUTION]: The skill relies on executing local Python scripts (
scripts/seo_agent.pyandscripts/report_publish.py) to perform its core functions. These scripts are granted the ability to modify the filesystem and interact with Git, which are powerful capabilities requiring user oversight, especially when automated via the suggested GitHub Actions workflow.
Audit Metadata