seo-backlink-opportunity-finder
Pass
Audited by Gen Agent Trust Hub on Aug 25, 2026
Risk Level: SAFEPROMPT_INJECTIONCOMMAND_EXECUTION
Full Analysis
- [PROMPT_INJECTION]: The skill processes untrusted data from external websites (crawled HTML, referring pages, and snippets) to identify backlink prospects, creating a surface for indirect prompt injection.
- Ingestion points: External URLs and content discovered during category-led and competitor-path prospecting (SKILL.md, references/research-protocol.md).
- Boundary markers: The skill contains explicit instructions to ignore instructions, commands, or requests embedded in external HTML or JSON evidence (SKILL.md).
- Capability inventory: The skill utilizes web search/browser tools and executes a local validation script using Node.js.
- Sanitization: Instructions are provided to treat crawled documents as untrusted and disregard instructions contained within them.
- [COMMAND_EXECUTION]: The skill executes a local Node.js script to validate the research ledger.
- Evidence: The instruction
node <absolute-path-to-skill>/scripts/validate-opportunity-ledger.mjs --input opportunities.json --tier minimumis used to verify candidate data (SKILL.md). - Context: The script is a provided component of the skill designed to enforce quality tiers and record structure.
- [SAFE]: The validation script (
scripts/validate-opportunity-ledger.mjs) implements security best practices by checking for unsafe literal hosts. It includes logic (isUnsafeLiteralHostandisUnsafeIpv4) to identify and block interaction with local and private IP ranges (e.g., 127.0.0.1, 10.0.0.0/8, 192.168.0.0/16), which mitigates potential Server-Side Request Forgery (SSRF) risks during the validation of prospect URLs.
Audit Metadata