seo-profound
Pass
Audited by Gen Agent Trust Hub on Oct 22, 2026
Risk Level: SAFECOMMAND_EXECUTIONDATA_EXFILTRATIONINDIRECT_PROMPT_INJECTION
Full Analysis
- [COMMAND_EXECUTION]: The skill directs the user to run local installation scripts (
extensions/profound/install.shandinstall.ps1). This is a standard setup procedure for the vendor's tools. - [DATA_EXFILTRATION]: The agent is instructed to verify a key in
~/.claude/settings.json. While used for configuration, reading the agent's settings file is a sensitive operation as it can contain keys for other services. - [INDIRECT_PROMPT_INJECTION]: The skill fetches data from external APIs and search results, creating a risk of indirect prompt injection. The skill includes defensive instructions to treat this data as untrusted.
- Ingestion points: API responses and fetched search results.
- Boundary markers: Included 'Security' section warns against treating fetched content as instructions.
- Capability inventory: SEO metrics retrieval via tool routing.
- Sanitization: Explicit instructions to report, not execute, any commands found in fetched content.
Audit Metadata