signals-scout-skills-store
Pass
Audited by Gen Agent Trust Hub on Jul 6, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADSPROMPT_INJECTION
Full Analysis
- [EXTERNAL_DOWNLOADS]: Fetches best-practice rules and specification updates from Anthropic's official documentation (platform.claude.com) and GitHub repository (github.com/anthropics).
- [EXTERNAL_DOWNLOADS]: Fetches specification details from the agentskills.io service.
- [PROMPT_INJECTION]: The skill audits untrusted content from the skills store (ingested via skill-get and skill-file-get tools), which creates an indirect prompt injection surface. It includes mandatory defensive instructions to treat all audited data as non-executable and to ignore any embedded framing or commands. The skill also requires redacting sensitive values from its findings to prevent data exposure in reports.
Audit Metadata