[KongchangAI]

#MCP

506 related articles

How Angular Uses MCP to Keep AI Coding Agents Up to Date
·2 min

How Angular Uses MCP to Keep AI Coding Agents Up to Date

Angular CLI integrates an MCP server from v19 onward, letting AI agents like Claude Code and Codex retrieve up-to-date docs via RAG to avoid training data cutoff issues.

Read more →
8 MCP Servers Worth Installing: Unlocking New Capabilities for AI Agents
·2 min

8 MCP Servers Worth Installing: Unlocking New Capabilities for AI Agents

Discover 8 free, open-source MCP servers — including dots for anti-detection browsing, Universal Modder for game mods, Concat for local video editing, and OpenSEO for real SEO data — that unlock new AI Agent capabilities.

Read more →
AWS MCP Server Expands to 6 New Regions: Infrastructure Acceleration for AI Coding Agents
·1 min

AWS MCP Server Expands to 6 New Regions: Infrastructure Acceleration for AI Coding Agents

AWS expands its managed MCP server to Singapore, Sydney, Tokyo, Ireland, London, and Oregon — reducing latency and meeting data residency requirements for AI coding agents.

Read more →
The AI Agent Payment Problem: Why LLMs Shouldn't Hold API Keys Directly
·2 min

The AI Agent Payment Problem: Why LLMs Shouldn't Hold API Keys Directly

Why giving AI agents direct payment API keys fails in production. Explore the three-layer risk of the Agent Execution Wall and an MCP-based propose-approve solution that zeroes out blast radius.

Read more →
context-mode: The Context Window Optimizer Built for AI Coding Agents
·2 min

context-mode: The Context Window Optimizer Built for AI Coding Agents

context-mode is an open-source context window optimizer for AI coding agents, offering 98% tool output reduction, persistent session memory, and MCP-based routing across 17 platforms.

Read more →
We've Been Using MCP Wrong: Its Only Real Value Is Authentication
·1 min

We've Been Using MCP Wrong: Its Only Real Value Is Authentication

A developer argues we're misusing MCP: its real value is on-demand authentication and permission isolation, not capability delivery. Here's when MCP actually makes sense.

Read more →
Adjudicated Query Pattern: Batch Lease Compliance Verification with Amazon Quick
·2 min

Adjudicated Query Pattern: Batch Lease Compliance Verification with Amazon Quick

AWS's Adjudicated Query pattern combines Amazon Quick, a bounded MCP server, and a deterministic rules engine to deliver provably complete, defensible compliance Q&A at scale.

Read more →
Moka Chat: One Command to Add MCP Tools to Your Ollama Models
·1 min

Moka Chat: One Command to Add MCP Tools to Your Ollama Models

Moka Chat is an MIT-licensed, local-first chat app. One npx command connects Ollama models to MCP tools, with Inspector showing JSON-RPC, token usage, and latency per call. Also supports LM Studio and hosted models.

Read more →
Pi 1.0 Release Breakdown: MCP, Codemode, and Pi Durable
·1 min

Pi 1.0 Release Breakdown: MCP, Codemode, and Pi Durable

Pi 1.0 is now out with built-in MCP via Codemode, a default full-screen TUI, and the new Pi Durable persistence mechanism. Here's what changed and what to watch out for when upgrading.

Read more →
MCP from Zero to One: A Complete Guide to Building Production-Grade MCP Servers
·2 min

MCP from Zero to One: A Complete Guide to Building Production-Grade MCP Servers

From local test scripts to production-grade MCP servers: a full guide covering MCP architecture, Streamable HTTP, stateless design, Docker, Nginx, and Kubernetes.

Read more →
When an AI Agent Truly Controls Your Homelab: Lessons and Risks from a K3s Experiment
·1 min

When an AI Agent Truly Controls Your Homelab: Lessons and Risks from a K3s Experiment

A developer connected a local Qwen Agent to a K3s homelab via MCP with full write permissions. This experiment log reveals the real risks and lessons of giving AI agents true system access.

Read more →
MCP Servers: From Single-User to Concurrent Production — Where Tutorials Leave You Stranded
·2 min

MCP Servers: From Single-User to Concurrent Production — Where Tutorials Leave You Stranded

MCP server works fine solo but times out with two users? We break down the concurrency, connection pool, and session isolation issues that tutorials never cover.

Read more →
MCP Gateway Showdown: Cloudflare vs. Okta vs. Auth0 vs. Microsoft
·2 min

MCP Gateway Showdown: Cloudflare vs. Okta vs. Auth0 vs. Microsoft

Cloudflare, Okta, Auth0, and Microsoft all offer MCP gateways. This comparison covers their auth, logging, and positioning differences, plus what your MCP server still needs to handle itself.

Read more →
freddy.health: Talk to Your Health Data in Plain Language with AI
·1 min

freddy.health: Talk to Your Health Data in Plain Language with AI

freddy.health connects Garmin, Oura, WHOOP and more to Claude and ChatGPT via MCP protocol, letting users query sleep, HRV, recovery and training data in natural language.

Read more →
Andrew Ng Teams Up with Anthropic: A Practical Guide to the Claude Code Course
·2 min

Andrew Ng Teams Up with Anthropic: A Practical Guide to the Claude Code Course

Andrew Ng and Anthropic's Claude Code course covers context management, MCP servers, parallel sessions, and Git Worktrees through three real-world projects — from RAG chatbot to Figma frontend.

Read more →
Pi.dev Says No to MCP: An Alternative Path for AI Tool Integration
·2 min

Pi.dev Says No to MCP: An Alternative Path for AI Tool Integration

Pi.dev's post "You Said No MCP" sparked debate on Hacker News. We analyze why they're skeptical of MCP and explore pragmatic strategies for AI tool integration.

Read more →
Deploy a WorkOS MCP Server with One Command: Why Configured ≠ Authenticated
·1 min

Deploy a WorkOS MCP Server with One Command: Why Configured ≠ Authenticated

WorkOS's workos mcp install command configures MCP servers for Claude Code, Codex, and Cursor in one step. Learn why configured ≠ authenticated and how to debug multi-agent failures.

Read more →
AI Agent Development: From Zero to Enterprise-Grade Deployment
·4 min

AI Agent Development: From Zero to Enterprise-Grade Deployment

A practical breakdown of AI Agent development: framework selection, MCP tool calling, RAG data preparation, and production evaluation — what actually matters for enterprise deployment.

Read more →