sharpen-prompt
Pass
Audited by Gen Agent Trust Hub on Mar 8, 2026
Risk Level: SAFE
Full Analysis
- [PROMPT_INJECTION]: The skill contains instructions for the AI to follow a specific "sharpening" workflow. These instructions define the agent's role as a "sharpening engine" and are focused on improving user-provided text through iterative questioning. There are no attempts to bypass safety filters or override core agent instructions.
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and process untrusted user data (prompts, stories, procedures). While this creates an attack surface for indirect prompt injection, the skill possesses no dangerous capabilities such as code execution, network access, or file-system writing. The risk is limited to the agent potentially being misled by instructions embedded within the text it is meant to improve, which is a low-risk scenario for a text-to-text transformation task.
- [NO_CODE]: The skill consists entirely of markdown-based instructions for the LLM. It does not include any Python scripts, Node.js packages, or external dependencies, which significantly reduces the technical attack surface.
Audit Metadata