parasite-and-pr

Pass

Audited by Gen Agent Trust Hub on Nov 12, 2026

Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADS
Full Analysis
  • [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest data from external sources such as landing pages, pricing pages, and blog posts to generate marketing content. This creates an attack surface where instructions hidden on those pages could influence the agent's output files.- Ingestion points: Processes content from external URLs (landing page, /pricing, canonical blog post) and local files (seo/kb/brand-ssot.md).- Boundary markers: Absent. The skill does not instruct the agent to use specific delimiters or to disregard instructions embedded within the ingested external data.- Capability inventory: The skill possesses file system write access for creating distribution bundles in the 'marketing/distribution/' directory.- Sanitization: Absent. While the skill enforces stylistic rules for the final prose, it lacks mechanisms to sanitize or validate the source content against adversarial instructions.- [EXTERNAL_DOWNLOADS]: The skill directs the agent to interact with and utilize external tool configurations and services hosted on non-whitelisted domains.- Evidence: Instructs the use of an MCP tool endpoint at 'https://aiseotracker.com/api/agent/mcp' for SEO tracking.- Evidence: References an external skill configuration file at 'https://linkdr.com/SKILL.md' for link building services.
Audit Metadata
Risk Level
SAFE
Analyzed
Nov 12, 2026, 03:55 AM