marketingskills-agent-skills

Pass

Audited by Gen Agent Trust Hub on May 16, 2026

Risk Level: SAFEEXTERNAL_DOWNLOADSCOMMAND_EXECUTIONPROMPT_INJECTION
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill instructions include commands to download marketing-related markdown files and templates from a public GitHub repository using curl and git clone. These resources are used to populate a local directory for agent reference.
  • [COMMAND_EXECUTION]: The installation guide suggests the use of standard shell commands such as mkdir, npx skills, and chmod to initialize the project environment. These are common practices for deploying development tools.
  • [PROMPT_INJECTION]: The skill architecture relies on the creation and consumption of local context files (e.g., product-marketing-context.md) which ingest user-supplied project details. While this is an intended feature for personalization, it establishes a potential surface for indirect prompt injection if the processed data contains instructions designed to manipulate agent behavior.
  • Ingestion points: .skills/product-marketing-context.md, .skills/competitor-analysis.md, .skills/customer-research.md.
  • Boundary markers: None defined in the skill documentation.
  • Capability inventory: File system modifications and network operations via standard CLI tools.
  • Sanitization: No specific sanitization or input validation mechanisms are described.
Audit Metadata
Risk Level
SAFE
Analyzed
May 16, 2026, 01:52 PM
Security Audit — agent-trust-hub — marketingskills-agent-skills