prompt-engeneering
Pass
Audited by Gen Agent Trust Hub on Mar 4, 2026
Risk Level: SAFENO_CODE
Full Analysis
- [SAFE]: The skill consists entirely of documentation in Markdown format. There are no executable scripts (.py, .js, .sh), binaries, or configuration files that trigger automated system actions. All technical content is provided for educational and reference purposes only.
- [NO_CODE]: No functional code or automation scripts are shipped within the skill package. The skill serves as a static knowledge base for agents to reference when asked about prompt engineering techniques.
- [PROMPT_INJECTION]: The reference documentation includes examples of adversarial prompts and injection techniques (such as 'Ignore previous instructions'). These are explicitly presented as research material and failure cases within a 'Failure Taxonomy' and are not used to direct the host agent to bypass its safety filters.
- [EXTERNAL_DOWNLOADS]: The README provides an installation command using 'npx', which downloads the 'skills' utility from the npm registry. This is a standard installation pattern for this ecosystem and targets a well-known service (npm).
Audit Metadata