Articles tagged “mcp”
64 articles
How to Build an Agent That Writes Code Instead of Calling Tools
One JSON tool call per turn is slow, token-heavy, and brittle. A short script that orchestrates tools in a sandbox is faster and cheaper. Here's how it works.
MCP now uses round-trips, not streams, for user input
The July 28 spec deprecated elicitation and sampling. Multi Round-Trip Requests (SEP-2322) replace both with a stateless pattern: the server returns InputRequiredResult, the client collects input, the client re-submits. Here's how to migrate.
9 in 10 MCP Servers Skip OAuth. The July 28 Spec Just Landed.
Only 8.5% of MCP servers use OAuth. The July 28 spec landed, and it doesn't make auth mandatory. Here's what it actually requires and what to fix first.
Your MCP tool descriptions are failing your agent
97% of real-world MCP tool descriptions contain quality problems that hurt agent accuracy and inflate token costs. Here's what the research found and how to fix it.
A2A and MCP: the agent protocol stack explained
MCP connects agents to tools. A2A connects agents to other agents. Together they form the full protocol stack for multi-agent CX systems. Here's how they work and why you need both.
Four protocols that power every AI agent
MCP handles tools, A2A handles agent delegation, ACP routes across enterprise frameworks, and ANP crosses org boundaries. Here's how the four protocols divide the work and when each belongs in your stack.
MCP's 2026 spec changes what your tools can do
The MCP spec RC is out with Extensions, Tasks, MCP Apps, and OAuth hardening. Here's what changes for CX agent builders before the July 28 final.
Building MCP tools your CX agents actually use
Most CX agents have too many poorly-described tools. Here's how to design and build MCP tools that agents call correctly — covering CRM connectors, knowledge retrieval, and action tools.
What the July 28 MCP spec means for your stateful agents
The July 28 MCP spec is the largest protocol revision since launch. Sessions are removed. Stateless request handling replaces the initialization handshake. Here's what breaks and how to fix it before the deadline.
Static API keys don't work for autonomous agents
When you hand an autonomous agent a static API key, you're giving it a skeleton key with no expiry. Here's the per-call permission model that replaces it.
Your AI agents need their own identity
Most CX agents share a single API key with every other service on your stack. Here's how to implement proper agent identity: OAuth 2.0, scope minimization, JIT provisioning, and instant kill switches.
Parallel tool calls: the 5x speedup in your agent
Sequential tool calls compound latency for no reason. Here's how to fan out independent tool calls in parallel and cut agent response time by up to 5x in production.
MCP tasks: how async tool calls fix your agent's timeout problem
The November 2025 MCP spec introduced Tasks: a call-now, fetch-later primitive that lets agents kick off slow operations without blocking. Here's how it works and how to build with it.
The MCP registry: managing a fleet of agent tools
When you have more than 3 MCP servers, you need a registry. Here's the architecture Pinterest and Cloudflare use to manage fleets of domain-specific tool servers at scale.
Managed Agents in 2026: Three Runtimes, Three Trade-Offs
Google, Anthropic, and OpenAI all shipped 'managed agents' in May 2026, and they mean completely different things. Here's what each runtime trades away for CX teams.
How to Use MCP Sampling, Roots, and Elicitation in CX Agents
Most MCP tutorials cover only server-side features: tools, resources, prompts. The three client capabilities (Sampling, Roots, Elicitation) unlock human-in-the-loop patterns that server tools alone can't. Here's how to use them.
How to Migrate Your MCP Server to Stateless Mode
The MCP 2026 release candidate makes stateless the recommended default. Your MCP server can now scale behind any load balancer without sticky routing. Here's how to migrate and use the new Tasks extension for async CX work.
MCP Without a Gateway Is a Production Liability
Raw MCP is great for prototyping. But production agents need audit trails, per-user identity, tool-level RBAC, and rate limiting, none of which the spec provides. Here's the gateway pattern that fills the gap.
How to Build an MCP Apps Widget for Your CX Agent
MCP Apps (SEP-1865) lets your MCP server deliver interactive forms, dashboards, and buttons inside AI chat conversations. Here's how to build one for CX agents.
How AG-UI Connects AI Agents to Any Frontend
AG-UI is the open event protocol that connects AI agent backends to any frontend. Here's how it works, why the protocol stack now has three layers, and how to wire it into a real CX agent.
MCP Webhooks: Build Event-Driven Agents That React in Real Time
MCP's request-response model breaks when agents need to react to external events. Here's how to build event-driven agents today using stateless HTTP plus webhooks, and what the June 2026 spec will make native.
How MCP Tool Descriptions Break Your Agent
New research shows 97% of MCP tool descriptions have quality issues that hurt agent accuracy. Here's what the smells look like, why they matter, and how to fix them.
AWS just gave your agent 15,000 cloud tools
The AWS MCP Server is now GA. One tool call reaches any of 15,000+ AWS APIs, sandboxed Python execution lets agents run multi-step operations, and Agent Skills replace heavyweight SOPs with on-demand guidance. Here's what changed and how to wire it.
MCP Apps: Build UIs That Render Inside AI Chat
MCP Apps let your tools return interactive HTML dashboards, forms, and visualizations that render inline in Claude, ChatGPT, and VS Code. Here's how to build them for CX agents.
AG-UI: The Protocol That Connects Agents to UIs
AG-UI is the open event-based protocol that streams AI agent state to any frontend in real time. Here's how it works, what events it defines, and how to wire it up in TypeScript.
MCP Auth in Production: Scopes, Tokens, and Tenant Isolation
Most MCP servers ship with no auth. Here's how to add OAuth 2.0 scopes, per-tenant tool sets, and client isolation before your MCP server becomes load-bearing production infrastructure.
Past 50 tools, function-calling accuracy falls off a cliff
Past 50 tools, function-calling accuracy falls off a cliff. Measure the curve on your own agent and recover accuracy with per-turn toolset scoping.
MCP tool description drift: the silent failure nobody alerts on
Edit an MCP tool description for clarity, lose 8% routing accuracy, and the eval suite stays green. How to detect, gate, and roll back the drift.
Stop Loading All Your MCP Tools at Once
Loading 50 MCP tools burns 72K tokens before your agent says a word. Progressive tool discovery fixes that: smaller context, sharper decisions, real code patterns.
When Your Customer's Browser Agent Resolves the Ticket Before Yours Picks Up
Browser Use, Operator, Mariner, Computer Use. Here's what happens to CX volumes when the customer's browser agent resolves the ticket before yours picks up.
When to Use a Supervisor, When to Let Agents Swarm
Supervisor burns 20-40% more tokens per run. Swarm hits a quality cliff past 8-10 handoffs. Start supervisor, graduate to swarm when latency bites.
MCP SSE Is Deprecated. Here's How to Migrate
SSE transport is being deprecated across major MCP platforms in 2026. Here's a practical migration guide from HTTP+SSE to Streamable HTTP, with TypeScript examples and a phased rollout strategy.
7 FastMCP mistakes that break your agent in production
FastMCP servers that work locally often fail at scale. Seven common mistakes, from missing annotations to monolithic tool sets, and how to fix each one.
MCP Is Now Open Infrastructure: Build for What's Next
MCP was donated to the Linux Foundation and the AAIF just held its first summit. What does the protocol becoming open infrastructure mean for what you build on top of it?
Your MCP server is a monolith. Here's how to fix it
MCP servers dump every tool into the context window, burning tokens before your agent reasons. Four patterns to fix it: decompose, filter, gateway, facade.
The no-code ceiling: when agent builders hit production
Visual agent builders get you to 80% fast. The last 20%, telephony, monitoring, testing, and memory, requires infrastructure they never intended to provide.
MCP Servers in Production: Observability from Day One
Instrument your MCP servers with OpenTelemetry for production-grade observability. Covers tracing tool calls, detecting loops, cost attribution, and alerting.
Build the MCP + A2A agent protocol stack from scratch
Wire an MCP server to an A2A agent that delegates tasks and calls tools. TypeScript and Python examples, Streamable HTTP transport, Agent Cards, and auth.
Your AI Agent's Context Window Is Already Half Full
System prompts, tool schemas, MCP descriptions, memory injection, conversation history. They all eat tokens before the user says a word. Learn where your context budget goes and how to manage it.
MCP vs A2A: Tools Protocol, Agents Protocol, and Why You Need Both
MCP connects agents to tools. A2A connects agents to each other. Most developers confuse them. This guide breaks down both protocols with architecture diagrams, real code, and a decision framework for production systems.
The Insurance Agent That Never Misquotes a Policy
How regional insurers deploy AI agents that answer policy questions accurately, intake claims end-to-end, and produce the audit trail regulators demand.
Build a Restaurant AI That Remembers Every Regular
Build an AI phone agent for a local restaurant that takes orders, answers menu questions, and remembers regulars. A developer side hustle worth $400/month per client.
50 Tools, Zero Memory. The Biggest Gap in AI Agents Today
AI agents can call 50 APIs but can't remember what you said yesterday. The tool layer is years ahead of the memory layer, and customers are paying the price.
Why MCP Exists: Tool Calling Shouldn't Need Adapter Code
OpenAI, Anthropic, and Google all implement function calling differently. MCP is emerging as the standard that saves developers from writing adapter code for every provider.
From Keyword Search to Shopping Memory
Build the intelligence layer for an AI shopping assistant: semantic product search with Commerce MCP, customer memory that persists across visits, and MCP tool registration for multi-channel deployment.
Why AI Shopping Still Feels Like a Search Bar
Most AI shopping assistants return walls of text. Learn how ChatKit widgets and Vercel AI SDK structured output turn AI recommendations into interactive product cards with images, prices, and add-to-cart buttons.
Context Engineering Is What Your Agent Actually Needs
Prompt engineering hits a wall with production AI agents. Context engineering fixes it. Build a full context pipeline with memory, RAG, history compression, and tool resolution.
The Three Protocols Every AI Agent Will Speak
The AI agent protocol stack has three layers: MCP for tools, A2A for agent-to-agent communication, and WebMCP for browser interaction. A practitioner's guide to how they work together in production.
Your AI Agent Has No Guardrails
Air Canada honored a refund its chatbot hallucinated. DPD's bot cursed at customers on camera. One e-commerce agent approved $2.3M in unauthorized refunds at 2:47 AM. Here is the five-layer guardrail architecture that prevents all three.
Every Tool Is an Injection Surface
Prompt injection moved from chat to tool calls. Anthropic, OpenAI, and Arcjet shipped defenses in the same month. Here's what changed, what works, and what your agent architecture needs now.
Why Browser Agents Waste 89% of Their Tokens
Browser agents burn 1,500-2,000 tokens per screenshot. Chrome 146's navigator.modelContext API lets websites expose structured tools instead, cutting token usage by 89% and raising task accuracy to 98%. Here's how WebMCP works.
Part 1: Claude's 7 Extension Points — The Mental Model
CLAUDE.md, Skills, Hooks, MCP Servers, Connectors, Claude Apps, Plugins — Claude's extension ecosystem is powerful but confusing. Here's the mental model that makes sense of all 7.
Part 3: MCP Servers vs. Connectors vs. Apps
All Claude Apps are Connectors. All Connectors are MCP Servers. Understanding this hierarchy — and when to build vs. use managed integrations — saves weeks of unnecessary engineering.
Part 4: All 7 Extension Points in One Production Codebase
50+ skills, multiple MCP servers, scoped rules, safety hooks — here's how all 7 Claude extension points compose in a real NestJS monorepo with 17 projects. What works, what fights, and what we'd do differently.
MCP Is Now the Industry Standard for AI Agent Integrations. Here's What That Means
MCP standardizes how AI agents connect to tools and data, replacing fragile, proprietary integrations with a universal protocol. Here's what it means for your agents.
Your Voice Agent Forgets Everything. Here's How to Fix That
How to add persistent memory, tools, and knowledge to Pipecat and LiveKit voice agents using the Chanl Python SDK — one SDK instead of assembling five services.
71% of organizations aren't prepared to secure their AI agents' tools
MCP gives AI agents autonomous access to real systems — and introduces attack vectors that traditional security can't see. A technical breakdown of tool poisoning, rug pulls, cross-server shadowing, and the defense framework production teams need now.
MCP Streamable HTTP: The Transport Layer That Makes AI Agents Production-Ready
MCP's Streamable HTTP transport replaced the original SSE transport to fix critical production gaps. This guide covers what changed, why it matters, and how to implement it in TypeScript with code examples.
Your agent has 30 tools and no idea when to use them
MCP tools give agents external capabilities. Skills give agents behavioral expertise. Learn the architecture of both, build them in TypeScript, and understand when to use each — and when you need both.
AI Agent Tools: MCP, OpenAPI, and Tool Management That Actually Scales
How production AI agents discover, execute, and manage tools — from MCP protocol to OpenAPI auto-importing, security sandboxing, and multi-tenant tool infrastructure.
Build your own AI agent tool system — what breaks when you add the 20th tool?
Build a complete tool system for customer-facing AI agents from scratch — registry, execution, auth, monitoring. Then learn what breaks when real customers start calling.
MCP Deep Dive: Advanced Patterns for Agent Tool Integration
Production MCP patterns for teams who've built their first server and need to scale it — OAuth 2.1 with PKCE, Streamable HTTP transport, gateways, sampling, dynamic tool registration, and multi-tenant security.
MCP Explained: Build Your First MCP Server in TypeScript and Python
Build a working MCP server from scratch in TypeScript and Python. Hands-on tutorial covering tools, resources, transports, and testing.
The MCP Marketplace Problem: Why Standardized Integrations Need Standardized Testing
5,800+ MCP servers, 43% with injection flaws. Standardized protocol doesn't mean standardized quality. Why every MCP integration needs automated testing.
Learn Agentic AI
Weekly. Patterns for shipping agents that work. MCP, scorecards, regression tests, prompts, model comparisons.
