seo-commerce
Pass
Audited by Gen Agent Trust Hub on Oct 16, 2026
Risk Level: SAFECOMMAND_EXECUTIONDYNAMIC_EXECUTIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands to validate audit results. The section 'Formal evidence handoff' in SKILL.md directs the agent to run a command using the pattern:
python "<suite-root>/scripts/validate_seo_findings.py" validate-findings <bundle>/seo-findings.json --bundle <bundle>. - [DYNAMIC_EXECUTION]: The skill uses dynamic paths to locate executable scripts at runtime. It instructs the agent to determine the
<suite-root>directory by reading a local configuration file (.seo-suite-runtime.json) or by using an environment variable (${CLAUDE_PLUGIN_ROOT}), and then executes a script from that location. - [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process data from external, untrusted sources such as merchant feeds, storefront pages, and product markup.
- Ingestion points: Captured visible product/category pages, raw/rendered markup, Google Merchant Center feeds, and Bing shopping feeds as specified in SKILL.md and commerce-evidence-protocol.md.
- Boundary markers: No explicit delimiters or instructions to ignore embedded instructions within the ingested content are defined in the provided files.
- Capability inventory: The skill allows the agent to write data to the local file system (
seo-findings.json) and execute Python-based CLI tools. - Sanitization: While the instructions emphasize data reconciliation and validation, there are no specific technical sanitization or escaping protocols defined for the content before it is processed by the agent or passed to the validation script.
Audit Metadata