seo-link-building

Pass

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONDATA_EXFILTRATIONCOMMAND_EXECUTION
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill identifies link building targets by processing external data sources (Google Search Console CSV exports) and scanning existing repository content. While this creates a potential attack surface for indirect prompt injection, the risk is mitigated by the skill's reliance on specialized scripts that perform literal, deterministic text edits with strict matching requirements rather than allowing the model to make unconstrained changes.
  • [DATA_EXFILTRATION]: The skill handles private performance data from Search Console. It implements explicit privacy guidelines, mandating that the agent must extract data into a scratch directory outside of the version-controlled repository and ensuring that sensitive metrics like clicks and impressions are never included in commit messages, PR descriptions, or issues.
  • [COMMAND_EXECUTION]: The skill executes local Node.js scripts to analyze link counts and apply text replacements. These scripts exclusively use standard Node.js built-in modules (e.g., node:fs, node:path) and do not download or execute untrusted third-party packages or remote code.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 12, 2026, 10:16 PM