seo-content
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill executes shell commands to run specific Python scripts for validating findings and managing markdown content, such as 'validate_seo_findings.py', 'manage_markdown_site.py', and 'manage_markdown_work.py'.
- [DYNAMIC_EXECUTION]: Script execution paths are dynamically constructed using variables like '' (extracted from '.seo-suite-runtime.json') or '${CLAUDE_PLUGIN_ROOT}', which determines the location of utility scripts at runtime.
- [INDIRECT_PROMPT_INJECTION]:
- Ingestion points: Untrusted external data enters the agent context through 'page/content capture' and 'visible page content' as specified in 'SKILL.md' and 'references/snippet-copy-protocol.md'.
- Boundary markers: The skill instructs the agent to require human fact/legal review and uses a 'claim ledger' to label data (e.g., 'verified fact', 'editorial interpretation') to distinguish source content from generated output.
- Capability inventory: The skill possesses the capability to execute shell-based Python scripts and interact with the file system to build and validate site artifacts, as described in 'SKILL.md' and 'references/markdown-delivery-contract.md'.
- Sanitization: There is no mention of automated programmatic sanitization of external content; the skill relies on manual human verification and the use of 'owner-approved sources'.
Audit Metadata