press-release-writing

Pass

Audited by Gen Agent Trust Hub on Apr 16, 2026

Risk Level: SAFE
Full Analysis
  • [EXTERNAL_DOWNLOADS]: The skill references installation documentation for its command-line tool at https://raw.githubusercontent.com/inference-sh/skills/refs/heads/main/cli-install.md. This is a legitimate documentation resource from the service provider.
  • [COMMAND_EXECUTION]: The skill utilizes the infsh CLI tool via the Bash tool to perform search tasks (e.g., infsh app run tavily/search-assistant). This is an expected and authorized use of the platform's research capabilities as defined in the allowed-tools configuration.
  • [PROMPT_INJECTION]: The skill presents a surface for indirect prompt injection as it processes user-provided press release details to perform research and generate content. However, this is inherent to the functionality of a writing assistant and no malicious instructions are present in the skill itself.
  • Ingestion points: User input defining the press release subject, quotes, and company background.
  • Boundary markers: The skill relies on natural language structure rather than explicit delimiters for user data.
  • Capability inventory: Search and research execution via the infsh CLI.
  • Sanitization: No specific sanitization logic is implemented, which is standard for instructions-based skills.
Audit Metadata
Risk Level
SAFE
Analyzed
Apr 16, 2026, 12:27 PM