Building Agents

Verify CopilotKit project wiring and scaffold new agent apps with the CLI

Create a Slack app and tokens to connect a CopilotKit Channels agent to Slack

Query a WeKnora knowledge base with chat, session ask or chunk search from the CLI

Drive a WeKnora RAG server from the CLI with auth, knowledge bases and JSON output

Red-team LLM apps in CI with Promptfoo and DeepTeam against OWASP LLM Top 10 presets

Add Llama Guard, NeMo Guardrails and LLM Guard filters to screen prompts and responses

Flag prompt injection in LLM inputs using regex, heuristics and a DeBERTa classifier

Scan web pages, PDFs and OCR text for hidden prompt injection before agents read them

Filter LLM inputs and outputs with NeMo Guardrails, PII checks and Guardrails AI

Run multi-turn adversarial tests on LLM apps with Microsoft PyRIT orchestrators

Lock down agent tool calls with allowlists, approvals, guardrails and audit logs

Test and benchmark LLM agents for behavior, reliability and production

Design agent tools with JSON schemas, clear descriptions and solid errors

Design autonomous AI agents with tool use, memory, planning and orchestration

Build AI product features with LLM patterns, RAG, prompt design and cost control

Call external models such as Grok, GPT and DALL-E for images and X data

Design and run evals for LLM, agent and retrieval systems with clear grading

Build with the Gemini API and Google Gen AI SDK in Python, JS, Go and more

Trace, evaluate and manage prompts for LLM apps with Langfuse

Apply production patterns for RAG pipelines, agents and LLM monitoring

Evaluate LLM apps with automated metrics, human review and A/B tests

Run LLM systems with RAG, embeddings, fine-tuning and quality evals

Cut chain-of-thought token use in reasoning models such as DeepSeek-R1 and QwQ

Look up current OpenAI API and SDK docs with citations while you build

Measure local model task latency, token use and errors with Ollama telemetry

Build typed AI agents with PydanticAI tools, structured output and DI

Design RAG systems with embeddings, vector stores and chunking strategies

Implement RAG chunking, embeddings and vector search for your app

Build low-latency voice agents that handle turns, interruptions and noise

Build voice AI apps with Realtime API, Vapi, Deepgram, ElevenLabs and LiveKit

Design and scaffold a new Nanobot agent from existing project tools and prompts

Set up 9Router, a local AI gateway with OpenAI-compatible APIs for chat and media

Run chat, code generation and streaming prompts through a 9Router OpenAI-style API

Generate embeddings through 9Router for RAG, semantic search and similarity

Triage large document corpora with cheap models and escalate top items to deep reads

Fetch LangGraph docs for stateful graphs, multi-step flows and approval steps

Playbook for building agents that write, review and refactor code

Playbook for building agents that draft blog posts, ads and marketing copy

Playbook for building agents that triage tickets, draft replies and escalate issues

Playbook for general-purpose personal assistant agents that fit no other archetype

Playbook for agents that run scheduled ops tasks, data syncs and notifications

Playbook for agents that search sources and write citation-backed research reports

Playbook for agents that read, update and report on spreadsheets and databases

Build, launch, grade and schedule a Claude Managed Agent in your own account

Build regression evals that prove an agent change did not make it worse

Challenge AI feature plans on evals, model choice, cost and risk before shipping

Choose API, fine-tune or in-house models and classify AI risk under the EU AI Act

Grade a managed agent against a rubric and iterate until a version passes evals

Interview a founder to scope a managed agent and draft its build sheet and eval plan