prompt-engineering-patterns
Pass
Audited by Gen Agent Trust Hub on May 19, 2026
Risk Level: SAFEEXTERNAL_DOWNLOADS
Full Analysis
- [SAFE]: The skill is an educational resource and toolkit for prompt engineering. All analyzed scripts and templates are consistent with this purpose, and no malicious patterns such as credential theft or unauthorized persistence were found.- [EXTERNAL_DOWNLOADS]: The skill references standard machine learning and LLM libraries including
openai,sentence-transformers,numpy,scipy, andscikit-learn. These are standard tools for the demonstrated functionality.- [PROMPT_INJECTION]: The skill contains patterns that interpolate untrusted data into prompts, which is characteristic of the prompt engineering domain but represents an indirect prompt injection surface. - Ingestion points: User-provided inputs (e.g.,
{query},{text}) are interpolated into templates inSKILL.md,references/prompt-templates.md, andassets/prompt-template-library.md. - Boundary markers: Absent; the provided examples rely on standard instructional formatting and delimiters without explicit hardening.
- Capability inventory: Code examples include network communication (via the
openailibrary inreferences/chain-of-thought.md) and local file writing (scripts/optimize-prompt.py). - Sanitization: No sanitization or escaping of user-provided content is implemented within the template rendering logic.
Audit Metadata