Get Agno SDK conceptual documentation and guides for writing agent code.

Args: path: Documentation path. Use "basics/" to see all topics.

SDK Documentation Paths: - "basics/agents/" - How to create and configure agents in code - "basics/tools/" - How to create custom tools for agents - "basics/memory/" - How to use memory in your agent code - "basics/knowledge/" - Knowledge bases and RAG implementation - "basics/teams/" - Multi-agent team coordination - "basics/workflows/" - Workflow orchestration patterns

This is for SDK/library usage (writing Python code with Agno). For deployed agent REST APIs and runtime features, use agno_agentos instead.

Get Agno SDK class and method reference (parameters, signatures, options).

Args: topic: Reference topic. One of: agents, teams, workflows, tools, models, memory, knowledge, storage, hooks, compression, reasoning, agent-os

This provides SDK CLASS documentation (e.g., Agent() constructor parameters, MemoryManager methods, Tool decorator options).

For runtime REST API endpoints (deployed agent APIs), use agno_agentos instead.

Get SDK code examples for building agents with Agno.

Args: category: Example category. One of: agents, teams, workflows, tools, memory, knowledge, models, database, evals, guardrails, hitl, multimodal, reasoning, sessions, tracing Leave empty to list all available categories.

Returns complete, runnable Python code examples with imports and setup. These are SDK examples for writing agent code, not deployment examples.

For deployment and hosting examples, use agno_agentos instead.

Get integration documentation for databases, vector stores, and models.

Args: integration_type: Type of integration. One of: database, vectordb, models, toolkits name: Specific integration name. Leave empty to list all of that type. Database: postgres, mongodb, sqlite, mysql, redis, dynamodb, firestore VectorDB: pinecone, qdrant, chroma, weaviate, milvus, lancedb Models: openai, anthropic, google, azure, bedrock

Examples: - integration_type="database", name="postgres" - integration_type="vectordb", name="pinecone" - integration_type="models" (lists all providers)

Get AgentOS runtime and deployment documentation (REST APIs, endpoints, hosting).

Args: path: Path within AgentOS docs. Leave empty for overview.

AgentOS Sections: - "api/" - REST API authentication and endpoint usage - "features/" - Runtime features: memories, sessions, knowledge, tracing - "features/memories" - Memory API endpoints for deployed agents - "security/" - RBAC, authentication, authorization - "middleware/" - JWT auth, custom middleware - "interfaces/" - Slack, WhatsApp, A2A protocol integrations - "client/" - AgentOS Python client for calling deployed agents

USE THIS TOOL for questions about:

• Deployed agent REST API endpoints

• Memory/session/knowledge endpoints in production

• Authentication and security for hosted agents

• Runtime features and middleware

For SDK code usage (writing agents), use agno_docs or agno_reference instead.

Get migration guides and FAQ documentation.

Args: topic: Topic to fetch. Leave empty to list all available topics. Migration guides: v2-migration, workflows-migration, installation, changelog FAQ topics: environment, openai-key, structured-outputs, docker-connection, agentos-connection, rbac-auth, switching-models, tpm, tableplus

Use for upgrading Agno versions, installation issues, and common errors.

Get AgentOS REST API endpoints from the OpenAPI specification.

Args: resource: API resource to look up. One of: memory, agents, teams, workflows, sessions, knowledge, evals, traces, metrics, database, playground Leave empty to list all available API resources.

Returns detailed REST API endpoint documentation including:

• HTTP method and path (e.g., GET /memories, POST /memories)

• Request parameters and their types

• Request body schema

• Response codes and descriptions

USE THIS TOOL when the user asks about:

• REST API endpoints for deployed AgentOS

• HTTP methods for interacting with AgentOS

• API schemas and parameters

• How to call AgentOS programmatically via HTTP

For SDK/Python code (classes, methods), use agno_reference instead. For conceptual docs about features, use agno_agentos instead.