Building Agents

Scaffold a minimal local LangGraph agent in Python from the official quickstart

Scaffold a minimal local LangGraph agent in TypeScript from the official quickstart

Build and deploy a first Managed Deep Agent in LangSmith, step by step

Evaluate agent behavior with Autocontext scenarios and export reusable knowledge

Run scenarios through the Autocontext improvement loop to create reusable knowledge

Generate, coach and debug grid_ctf strategies with playbooks and lessons

Build and debug RL environments for Atropos training with reward functions and evals

Produce emotion and intention JSON files from workspace context

Fetch the current world observation for each simulation tick

Generate a new agent with its Python code and MCP config from requirements

Build voice and chat AI agents with LiveKit Agents, handoffs and tool workflows

Drive multi-turn conversations with a local LiveKit agent to see how it behaves

Look up current LiveKit APIs, CLI flags, model support and pricing from the docs

Build golden-dataset eval suites and regression gates for AI agent releases

Trace AI agents with token, latency and cost metrics for debugging

Lower LLM spend with model routing, batching, budgets and cost tracking

Front LLM APIs with a gateway for rate limits, key management, fallbacks and costs

Build LLMOps pipelines with model promotion, eval gates and rollbacks

Build RAG infrastructure with embeddings, chunking, hybrid search and reranking

Measure RAG retrieval quality, groundedness and hallucinations in regression tests

Write declarative LM programs in DSPy and auto-optimize prompts and RAG pipelines

Generate structured JSON, regex-constrained or Pydantic-typed output from LLMs

Join Google Meet or Zoom calls as a video agent with a configurable avatar

Scaffold a first eval suite that turns real agent failures into regression cases

Find pipeline decisions that do not need the big model and route them to cheaper gates

Build LLM-powered features with prompt engineering, RAG and chatbot patterns

Add AI text and image generation to Node.js backends with the CloudBase SDK

Call text and streaming AI models directly from browser apps with the CloudBase JS SDK

Add text generation and streaming AI to WeChat mini programs with wx.cloud

Build and deploy AI agent servers with the CloudBase SDK and AG-UI streaming

Build AI agent backends in Python with LangGraph, FastAPI and AG-UI streaming

Diagram an Agentforce agent's topics, actions and flows from its metadata

Convert Einstein Bots into Agentforce agents in one guided pass

Trace one Agentforce session end to end using Data Cloud records

Build and debug Agentforce agents written in Agent Script .agent files

Analyze production Agentforce traces and manage health monitoring alerts

Replay past Agentforce conversations against a different LLM to compare

Write and run functional and security test suites for Agentforce agents

Set up the Education Cloud Student Recruitment Agent for admissions questions

Create Custom Lightning Types with JSON schemas for agent actions

Turn Agentforce execution trace spans on or off for Data Cloud

Connect an active Agentforce agent to a messaging channel with routing

Set up an Agentforce Contact Center voice channel with a live phone number

Configure handoff from an Agentforce agent to a staffed human queue

Set up and launch a Salesforce Help Agent with a guided checklist

Grant the permission sets an activated ITSM agent's actions need to run

Set up Agentforce for IT Service with fulfiller and employee agents in order

Turn on Agentforce Studio feature toggles for IT Service agents using sf CLI

Check Agentforce IT Service prerequisites for fulfiller and employee agents