vs-and-alternatives

Pass

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill instructs the agent to mine and process untrusted data from public sources including G2, Capterra, Trustpilot, GitHub Issues, and Reddit to identify "switching triggers" and competitor weaknesses.
  • Ingestion points: Automated reading/scraping of public review platforms and community forums (SKILL.md, Step 1).
  • Boundary markers: The instructions lack technical delimiters or boundary markers for the data being ingested; however, they require the agent to rewrite emotional claims into factual statements.
  • Capability inventory: The skill has the capability to write new MDX and TSX files to the user's repository and modify existing navigation/footer components.
  • Sanitization: The skill relies on verification gates (recurring triggers, current documentation checks) which act as a logical filter for external content.
  • [EXTERNAL_DOWNLOADS]: The skill encourages the integration of external Model Context Protocol (MCP) servers and additional skills from third-party domains to track and amplify the results of the generated content.
  • Evidence: References to https://aiseotracker.com/api/agent/mcp and https://linkdr.com/SKILL.md (SKILL.md, Track and amplify section).
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 12, 2026, 03:54 AM