> ## Documentation Index
> Fetch the complete documentation index at: https://docs.useagentflow.co/llms.txt
> Use this file to discover all available pages before exploring further.

# SDK Resource Index

> Complete namespace map for the Python and TypeScript AgentFlow clients

Both SDKs mount the resources below on the root client. Method names use `snake_case` in both languages unless a page documents a TypeScript alias or a deliberate cross-SDK difference.

## Agent runtime and configuration

| Namespace                  | Main operations                                                   | Guide                                        |
| -------------------------- | ----------------------------------------------------------------- | -------------------------------------------- |
| `client.agents`            | discover, configure, run, stream, manifests, context              | [Agents](/sdk/agents)                        |
| `client.agents.subagents`  | lifecycle, assignments, run, stream                               | [Agents](/sdk/agents)                        |
| `client.agents.prompts`    | Prompt Builder list, preview, validate, publish, reset            | [Prompt Blocks](/sdk/prompt-blocks)          |
| `client.agents.tools`      | create, register, configure, assign, run, dry-run                 | [Tools](/sdk/tools)                          |
| `client.agents.guardrails` | list, assign, detach                                              | [Guardrails](/sdk/guardrails)                |
| `client.prompt_blocks`     | reusable block CRUD, preview, cache invalidation, override revert | [Prompt Blocks](/sdk/prompt-blocks)          |
| `client.models`            | model catalog, provider configuration, key tests, allowlists      | [Models](/concepts/models)                   |
| `client.llm`               | chat, autocomplete, and model recommendation helpers              | [Models](/concepts/models)                   |
| `client.completions`       | run, classify, extract, summarize, and enrich convenience calls   | [Run Agent](/api-reference/agents/run-agent) |

## Context, data, and outputs

| Namespace                                | Main operations                                                | Guide                                                                               |
| ---------------------------------------- | -------------------------------------------------------------- | ----------------------------------------------------------------------------------- |
| `client.conversations`                   | create, list, messages, goals, markers, runs, watch            | [Conversations](/sdk/conversations)                                                 |
| `client.memory`                          | create, list, retrieve, update, recall, delete                 | [Memory](/sdk/memory)                                                               |
| `client.knowledge_bases`                 | create, ingest, search, documents, operations                  | [Knowledge Bases](/sdk/knowledge-bases)                                             |
| `client.documents`                       | safe document viewer URLs and content                          | [Documents](/sdk/documents)                                                         |
| `client.attachments`                     | upload, list, retrieve, content, download, delete              | [Attachments API](/api-reference/resources/attachments/upload-universal-attachment) |
| `client.cached_results` / `client.cache` | retrieve cached tool payloads and paginated items              | [Tools](/sdk/tools#cached-results)                                                  |
| `client.artifacts`                       | instances, content/state, actions, types, schemas, assignments | [Artifacts](/sdk/artifacts)                                                         |
| `client.skills`                          | catalog, content, import, administration, assignments          | [Skills](/sdk/skills)                                                               |
| `client.mention_types`                   | manifest, search, editor discovery, preview, administration    | [Mentions](/sdk/mentions)                                                           |
| `client.activity`                        | unified account/contact activity timeline                      | [Activity](/concepts/activity)                                                      |
| `client.wiki`                            | page list/tree/search plus page CRUD and move                  | [Wiki API](/api-reference/resources/wiki/list-wiki-pages-endpoint)                  |

## Extensions and integrations

| Namespace             | Main operations                                                   | Guide                                                           |
| --------------------- | ----------------------------------------------------------------- | --------------------------------------------------------------- |
| `client.integrations` | configure, test, install, remove native integrations              | [Integrations](/sdk/integrations)                               |
| `client.mcp_catalog`  | curated MCP list, retrieve, install                               | [MCP](/sdk/mcp)                                                 |
| `client.mcp_servers`  | servers, credentials, OAuth, test, discover, import               | [MCP](/sdk/mcp)                                                 |
| `client.toolkits`     | validate, import, list, export, uninstall                         | [Toolkits](/sdk/toolkits)                                       |
| `client.tools`        | tenant-global listing, approval/enabled state, permanent deletion | [Tools](/sdk/tools#global-catalog-toolkits-and-mcp-tools)       |
| `client.emails`       | draft, create draft, send, and query email workflows              | [Email API](/api-reference/resources/email/draft-email-with-ai) |
| `client.slack`        | Slack settings, OAuth, installations, health, and user links      | [Slack Runtime](/guides/slack-runtime)                          |
| `client.calls`        | retrieve a voice call or look it up by phone number               | [Voice API](/api-reference/resources/voice/get-call-status)     |

## Human interaction and durable work

| Namespace              | Main operations                                                     | Guide                                                                           |
| ---------------------- | ------------------------------------------------------------------- | ------------------------------------------------------------------------------- |
| `client.approvals`     | pending requests, decisions, cancellation, stats, bulk response     | [Approvals](/sdk/approvals)                                                     |
| `client.questions`     | respond to an agent question                                        | [Approvals](/sdk/approvals)                                                     |
| `client.batches`       | submit, wait, stream, output, retry, cancel, delete                 | [Batches](/sdk/batches)                                                         |
| `client.background`    | status, output/result/message, and cancellation for background work | [Batches](/concepts/batch-processing)                                           |
| `client.cancellations` | cancel calls, runs, messages, tasks, or all active work             | [Cancellation API](/api-reference/resources/cancellation/get-active-executions) |
| `client.tasks`         | task list, create, retrieve, update, toggle, delete                 | [Tasks API](/api-reference/resources/tasks/list-tasks)                          |
| `client.agent_lab`     | comparison targets, sessions, messages, and lane timelines          | [Agent Lab API](/api-reference/resources/agent-lab/list-agent-lab-targets)      |

## Tenant administration and system status

| Namespace         | Main operations                                        | Guide                                                         |
| ----------------- | ------------------------------------------------------ | ------------------------------------------------------------- |
| `client.settings` | user settings, guardrails, service tier, display names | [Settings](/sdk/settings)                                     |
| `client.brand`    | brand profile/assets read, update, regenerate, reset   | [Brand API](/api-reference/resources/brand/get-brand-profile) |
| `client.admin`    | usage, budgets, runs, queues, wakeups, audit events    | [Usage & Budgets](/guides/usage-and-budgets)                  |
| `client.system`   | health, readiness, and version                         | [System API](/api-reference/resources/system/health-check)    |

The TypeScript client additionally exposes `client.knowledge_base_operations` for direct ingestion-operation lookup and cancellation, `client.meta` for product metadata, and `client.recommendations` for the recommendation helper API. These three resources do not currently have Python counterparts. Python exposes knowledge-base operation methods through `client.knowledge_bases` instead.

System method names differ slightly: Python uses `client.system.ready()` and additionally exposes `client.system.features()`; TypeScript uses `client.system.readiness()`. Both expose `health()` and `version()`. `client.transport` is not part of the stable resource API.

<Info>
  Administrative mutations require the corresponding tenant-admin or machine scope. A resource being mounted on the client does not imply that every authenticated principal can use every method.
</Info>