Tools for going past the chat box
Official model documentation, prompt-engineering reading, and the open-source tooling that turns one-off prompts into repeatable workflows. For the annotated reading list on prompting itself, see the Source Shelf.
Official model documentation
Each vendor's primary reference — the place to check capabilities, limits, and prompting recommendations straight from the maker.
Anthropic docs
docs.anthropic.comAPI reference and prompting documentation for the reasoning-focused model family.
OpenAI API docs
platform.openai.com/docsAPI reference, guides, and the prompt engineering section for GPT-class models.
Google AI / Gemini docs
ai.google.devDeveloper docs for Gemini — the model behind our ask box — including multimodal prompting.
Cohere docs
docs.cohere.comEnterprise-oriented documentation: RAG, structured outputs, and grounding.
Prompt engineering reading
The shortest reliable reading list. The full annotated version lives on the Source Shelf.
Prompt Engineering Guide (DAIR.AI)
promptingguide.aiCommunity-maintained technique reference, tied back to the research.
Anthropic prompt engineering guide
docs.anthropic.comThe vendor's prompting playbook — roles, examples, chain-of-thought.
OpenAI prompt engineering guide
platform.openai.com/docsSix strategies with worked examples for GPT-class models.
Open source & tools
When a prompt becomes a workflow, these are the workbenches.
OpenAI Evals
github.com/openai/evalsFramework for evaluating model (and prompt) performance systematically.
PromptFoo
github.com/promptfoo/promptfooTest harness for prompts: compare variants across models before you ship one.
The Ask 'em AI way
Start at the desk with a real question, learn the patterns in the prompt guide, and reach for these tools when a prompt you keep retyping deserves to become a workflow.
Vetted & shelved