prompt-gen

Pass

Audited by Gen Agent Trust Hub on Jun 21, 2026

Risk Level: SAFE
Full Analysis
  • [SAFE]: The skill operates as a specialized assistant for prompt engineering, guiding the user through a structured process to refine AI instructions. Its behavior aligns with its stated purpose.
  • [COMMAND_EXECUTION]: The skill provides instructions for the agent to save configuration files to the local file system (e.g., in ~/.config/opencode/agents/). This is a functional feature that assists users in managing their agents and requires user interaction and explicit configuration parameters before execution.
  • [PROMPT_INJECTION]: The skill processes untrusted user input to generate prompt templates, which constitutes an indirect prompt injection surface.
  • Ingestion points: User input provided in Step 1 of the SKILL.md for task descriptions and optimization requests.
  • Boundary markers: Explicit delimiters for the user-provided data are not specified in the instruction set.
  • Capability inventory: The skill involves writing generated prompts to configuration files on the disk as described in Step 6.3 of SKILL.md.
  • Sanitization: The skill relies on its quality self-check workflow (Step 5) and structured templates to maintain prompt integrity rather than explicit data sanitization techniques.
Audit Metadata
Risk Level
SAFE
Analyzed
Jun 21, 2026, 07:38 AM
Security Audit — agent-trust-hub — prompt-gen