# Chapters

- [Hello World](https://primer.ought.org/chapters/hello-world.md): The simplest recipe
- [Question Answering](https://primer.ought.org/chapters/question-answering.md): The first recipe that calls a language model
- [Q\&A without context](https://primer.ought.org/chapters/question-answering/q-and-a-without-context.md): Answering questions without extra information
- [Q\&A about short texts](https://primer.ought.org/chapters/question-answering/q-and-a-about-short-texts.md): Answering questions about a few paragraphs
- [Debate](https://primer.ought.org/chapters/debate.md): A recipe that composes calls to multiple agents
- [Representing debates](https://primer.ought.org/chapters/debate/representing-debates.md): Debates have turns, turns have authors and messages
- [From debates to prompts](https://primer.ought.org/chapters/debate/from-debates-to-prompts.md): Rendering a debate to a string
- [The debate recipe](https://primer.ought.org/chapters/debate/the-debate-recipe.md): Going back and forth between agents
- [Long Texts](https://primer.ought.org/chapters/long-texts.md): Answering questions about research papers
- [Loading paper text](https://primer.ought.org/chapters/long-texts/loading-paper-text.md): Loading papers as structured data
- [Finding relevant paragraphs](https://primer.ought.org/chapters/long-texts/finding-relevant-paragraphs.md): Paragraph-wise classification with parallelism
- [Answering given paragraphs](https://primer.ought.org/chapters/long-texts/answering-given-paragraphs.md): Answering the question given the top paragraphs—with subrecipes!
- [Amplification](https://primer.ought.org/chapters/amplification.md): Recursive question-answering
- [Asking subquestions](https://primer.ought.org/chapters/amplification/asking-subquestions.md): From question to more questions
- [Answering subquestions](https://primer.ought.org/chapters/amplification/answering-subquestions.md): Mapping our answerer over subquestions
- [One-step amplification](https://primer.ought.org/chapters/amplification/one-step-amplification.md): Answering given subquestion answers
- [Recursive amplification](https://primer.ought.org/chapters/amplification/recursive-amplification.md): Subquestions can have subquestions
- [Verifiers](https://primer.ought.org/chapters/verifiers.md): Sometimes evaluation is easier than generation
- [Checking answers](https://primer.ought.org/chapters/verifiers/checking-answers.md): Does this sound plausible?
- [Checking reasoning steps](https://primer.ought.org/chapters/verifiers/checking-reasoning-steps.md): Where did we go wrong?
- [Tool Use](https://primer.ought.org/chapters/tool-use.md): Language models for controlling web search and Python interpreters
- [Web search](https://primer.ought.org/chapters/tool-use/web-search.md): Running web searches for getting current information
- [Interpreters](https://primer.ought.org/chapters/tool-use/interpreters.md): Executing code for more accurate computation
- [Deduction](https://primer.ought.org/chapters/deduction.md): Drawing conclusions from premises
- [Chain of Thought](https://primer.ought.org/chapters/deduction/chain-of-thought.md): Let’s think step by step.
- [Action Selection](https://primer.ought.org/chapters/action-selection.md): So many things a model could do
- [One-shot action selection](https://primer.ought.org/chapters/action-selection/one-shot-action-selection.md): Choosing a cognitive action
- [Iterative action selection](https://primer.ought.org/chapters/action-selection/iterative-action-selection.md): Choosing a sequence of actions
- [Amplification Revisited](https://primer.ought.org/chapters/amplification-revisited.md): Subquestions and so much more


---

# Agent Instructions: Querying This Documentation

If you need additional information that is not directly available in this page, you can query the documentation dynamically by asking a question.

Perform an HTTP GET request on the current page URL with the `ask` query parameter:

```
GET https://primer.ought.org/chapters.md?ask=<question>
```

The question should be specific, self-contained, and written in natural language.
The response will contain a direct answer to the question and relevant excerpts and sources from the documentation.

Use this mechanism when the answer is not explicitly present in the current page, you need clarification or additional context, or you want to retrieve related documentation sections.