Skip to content

Research Positioning

The repository was positioned after reviewing official provider documentation and reference projects such as DAIR Prompt Engineering Guide, prompts.chat, Anthropic courses, Learn Prompting, Promptfoo, OpenAI Cookbook evaluation examples, OWASP, NIST, MITRE ATLAS, and provider docs from OpenAI, Anthropic, Google, Microsoft, Meta, Mistral, and Amazon Bedrock.

Differentiation

  • Learn: concise framework and pattern explanations.
  • Choose: catalogs organized by task, provider, failure mode, and category.
  • Copy: reusable templates with variables and acceptance criteria.
  • Test: rubric, before-and-after examples, and validation scripts.
  • Operate: context engineering, tool provenance, security, and governance.

Existing resources are useful but often focus on education, large prompt lists, single-provider advice, or evaluation tools. This project connects those needs in a compact, machine-readable reference that can be checked locally.