# Docs - Docklee - **Getting Started** - [Quickstart](/docs/docklee/quickstart): Install the SDK, store your first memory, and get a cited answer from your own knowledge in a few minutes. - [Overview](/docs/docklee/overview): Give your AI app knowledge it can answer from and memory that lasts between sessions, through one API. - [API keys](/docs/docklee/api-keys): Create a key, choose what it can do and reach, send it with each request, and rotate it without downtime. - [Memory vs Knowledge](/docs/docklee/memory-vs-knowledge): Decide whether your data belongs in a knowledge engine, a memory space, or both. - **Core Concepts** - [How Docklee works](/docs/docklee/how-docklee-works): Follow data from a write, into a container or a memory space, and back out through Lynk. - [Sources](/docs/docklee/sources): Connect documents, apps, and data once, and use them in any engine. - [Containers](/docs/docklee/containers): Where knowledge is stored and processed, and how to declare a vector or OKF container on a write. - [Engines](/docs/docklee/engines): An engine reads from the sources linked to it and answers through Lynk. - [Memory spaces](/docs/docklee/memory-spaces): The container your memories live in, how to point the SDK at one, and how to keep spaces apart. - [End users](/docs/docklee/end-users): Keep each of your users' memories private inside one memory space. - **Support** - [FAQs](/docs/docklee/faqs): Answers to the questions developers ask most when they start with Docklee. - Memory - **Core Concepts** - [How memory works](/docs/memory/how-memory-works): What Docklee does with every memory you write, and how it decides what to return when you recall. - [Capture memory](/docs/memory/capture-memory): Write memories from your app with Dock, with a type, metadata, and an end user when the memory belongs to one person. - [Recall memory](/docs/memory/recall-memory): Ask a memory space a question with Lynk and get back the most relevant, current memories. - [Memory types](/docs/memory/memory-types): The type you send with a capture, the type Docklee assigns, and the storage class recall returns. - **Features** - Essentials - [Metadata](/docs/memory/essentials/metadata): Attach your own keys and values to a memory, such as a ticket ID or a channel. - [Duplicates](/docs/memory/essentials/duplicates): How Docklee handles a capture that repeats one it already has. - [Batching and flush](/docs/memory/essentials/batching-and-flush): When Dock sends your captures, and how to make sure none are lost when your process exits. - Intelligence - [Contradictions and updates](/docs/memory/intelligence/contradictions-and-updates): How a newer fact replaces an older one, and what happens when two facts can't both be true. - [Confidence](/docs/memory/intelligence/confidence): How much Docklee trusts each memory, and how that trust changes as facts are confirmed or challenged. - [Entities and links](/docs/memory/intelligence/entities-and-links): How Docklee finds the people, companies, and things in each memory, and connects memories to each other. - [Consolidation](/docs/memory/intelligence/consolidation): How Docklee folds old, low-value details into summaries while keeping important memories in full. - [Ranking](/docs/memory/intelligence/ranking): The signals Docklee combines to decide which memories a recall returns, and in what order. - Automatic Capture - [Instrument your app](/docs/memory/automatic-capture/instrument): Capture your web app's write requests as memories, with one line of setup. - [Watch a database](/docs/memory/automatic-capture/watch): Capture new and updated rows from your database tables as memories. - [Listen for webhooks](/docs/memory/automatic-capture/listen): Receive webhooks from the tools you already use and capture them as memories. - Governance - [Review and correction](/docs/memory/governance/review-and-correction): Check what Docklee stores, settle the conflicts it flags, and fix memories that are wrong. - [Retention](/docs/memory/governance/retention): Keep memories forever or for a set time, and keep chosen memories longer than the rest. - [Content Shield](/docs/memory/governance/content-shield): Hide what your memories say from people who use the dashboard, while your app keeps working as normal. - [Deleting a person's data](/docs/memory/governance/entity-deletion): Erase every memory about one person, company, or other entity in a memory space, permanently. - [Legal hold](/docs/memory/governance/legal-hold): Preserve every memory exactly as it is while a legal or compliance matter is open. - **Support** - [Troubleshooting](/docs/memory/troubleshooting): Fix the most common problems with capturing and recalling memories. - Knowledge - **Core Concepts** - [How knowledge works](/docs/knowledge/how-knowledge-works): How your content becomes searchable knowledge, and how an engine turns a question into a cited answer. - [Capture knowledge](/docs/knowledge/capture-knowledge): Write knowledge from your code into a vector or OKF container with Dock, and make it answerable by an engine. - [Query an engine](/docs/knowledge/query-an-engine): Ask a question and get one answer from your own content, with citations and a confidence score. - [Retrieve context](/docs/knowledge/retrieve-context): Get the most relevant chunks from an engine and give them to the model you already use. - [Streaming answers](/docs/knowledge/streaming-answers): Show an answer as the engine writes it, so people see the first words in moments. - **Features** - Ingestion - [Chunking](/docs/knowledge/ingestion/chunking): How a vector container splits documents into chunks that keep sections, paragraphs, and tables intact. - [Content screening](/docs/knowledge/ingestion/content-screening): How every write is checked for leaked secrets, injected instructions, and personal data before it's stored. - Answers - [Answer modes](/docs/knowledge/answers/answer-modes): Choose how strictly an engine's answers stay within your sources. - [Citations and confidence](/docs/knowledge/answers/citations-and-confidence): How each answer is checked against your sources, and how to use citations and confidence in your app. - [Answer cache](/docs/knowledge/answers/answer-cache): How a repeat of a recent question is answered in milliseconds from an encrypted cache. - Retrieval - [Vector and OKF together](/docs/knowledge/retrieval/vector-and-okf): Link vector and OKF sources to one engine, and it searches both and merges the results. - [Precision Mode](/docs/knowledge/retrieval/precision-mode): Index a container's sources more deeply, so common questions are answered faster and more precisely. - [Cross-document reasoning](/docs/knowledge/retrieval/cross-document-reasoning): How Docklee compares each new document with related ones, so answers know when documents agree, disagree, or replace each other. - **Support** - [Troubleshooting](/docs/knowledge/troubleshooting): Fix the most common problems with capturing knowledge, querying engines, and streaming answers. - OKF - **Core Concepts** - [What is OKF](/docs/okf/what-is-okf): Google defined a format for writing knowledge down. Docklee turns that format into retrieval your app can ask questions of. - [How OKF retrieval works](/docs/okf/how-okf-retrieval-works): How an engine finds the concepts a question is about, follows their links, and returns a small, ranked set of context. - [OKF retrieval vs vector RAG](/docs/okf/okf-vs-vector): Which questions each kind of retrieval answers best, and when to link both to one engine. - **Build OKF** - [Generate OKF](/docs/okf/generate-okf): How Docklee turns your documents, apps, and Dock writes into concepts and the links between them. - [Concept files](/docs/okf/concept-files): What every concept file holds, from the standard fields to the confidence, authority, and links Docklee adds. - [Semantic vocabulary](/docs/okf/semantic-vocabulary): The concept types and relationship types Docklee uses, so every concept and link means the same thing everywhere. - [Merging and conflicts](/docs/okf/merging-and-conflicts): How Docklee keeps one concept per idea, and what happens when new content disagrees with what's there. - **Use OKF** - [Query OKF](/docs/okf/query-okf): Link OKF sources to an engine and read them with the same query and retrieve calls you already use. - [Export OKF](/docs/okf/export-okf): Download your OKF bundle as plain Markdown files, readable by any tool that supports the format. - [OKF in memory](/docs/okf/okf-in-memory): How Docklee also turns a memory space into a concept graph, so recall can reason over concepts as well as single memories. - SDKs - **Getting Started** - [Overview](/docs/sdks/overview): One package for Python and TypeScript, with Dock to write and Lynk to read. - [Python quickstart](/docs/sdks/python-quickstart): Install the Python SDK, store a memory, recall it, and query an engine from Python in a few minutes. - [TypeScript quickstart](/docs/sdks/typescript-quickstart): Install the TypeScript SDK, store a memory, recall it, and query an engine from Node.js in a few minutes. - **Reference** - [Configuration](/docs/sdks/configuration): Set the SDK up with environment variables or constructor options, and control timeouts, retries, and the API address. - [Python reference](/docs/sdks/python): Every class and method in the Python SDK, with its options and what it returns. - [TypeScript reference](/docs/sdks/typescript): Every class and method in the TypeScript SDK, with its options and what it returns. - [Types](/docs/sdks/types): Every result the SDK returns, with its fields in Python and TypeScript. - [Errors and retries](/docs/sdks/errors-and-retries): Which failures the SDK retries for you, and which error it raises when a call can't succeed. - API Reference - **Getting Started** - [API overview](/docs/api/overview): REST endpoints for writing knowledge and memory, recalling memories, and asking engines. - [Errors](/docs/api/errors): The status codes the API returns, the shape of each error body, and the error codes you can match on. - [Rate limits](/docs/api/rate-limits): How many requests each API key can make, what happens when you hit a limit, and how to stay under it. - **Dock** - [Ingest events](/docs/api/ingest): Write a batch of up to 100 events to a knowledge container, a memory space, or both. - **Memory** - [Search memories](/docs/api/search): Recall the memories in one memory space that best answer a question. - [Write a memory](/docs/api/memory-write): Write one memory to a memory space and get its ID back. - [Search many memory spaces](/docs/api/federated-search): Recall from up to 10 memory spaces in one call, merged into one ranked list. - **Knowledge** - [Query an engine](/docs/api/query): Ask an engine a question and get one answer from your sources, with citations and a confidence score. - [Stream an answer](/docs/api/query-stream): Ask an engine a question and receive the answer as server-sent events while it's written. - [Retrieve chunks](/docs/api/retrieve): Get the ranked chunks an engine finds for a question, without writing an answer. - **Context** - [Assemble context](/docs/api/context-assemble): Get an engine's answer and the memories that relate to it in one call, and save the turn to memory.