Wiring MCP Servers Into Claude Cowork Safely
Connect MCP servers to Claude Cowork safely: scoped auth, tight tool schemas, structured error handling, and idempotency keys that prevent duplicate actions.
8 min read
Agentic AI & LLMs
Agentic AI & LLMs archive page 297 of 631
Agentic AI & LLMs
Agentic AI, LLM engineering, and the models behind modern automation — multi-agent systems, LLM evaluation and comparisons, RAG, fine-tuning, AI infrastructure, security, and production AI engineering.
From the blog
Latest articles
Reusable Patterns for Building Claude Cowork Agents
Code-level patterns for Claude Cowork: single-responsibility skills, typed tool contracts, layered context, intent-named idempotent tools, and clean handoffs.
7 min read
How to Deploy Claude Cowork: A Walkthrough
Step-by-step Claude Cowork deployment: build a read-only MCP connector, write and test a skill, bundle a role plugin, and pilot safely with one team.
8 min read
Claude Cowork Architecture: How the Pieces Fit (Deploy Cowork Across Enterprise)
End-to-end look at Claude Cowork's internals — plugins, skills, MCP connectors, and sub-agents — and how to govern each layer in an enterprise rollout.
8 min read
Computer Use 2.0 Goes GA: What Changed Since the Beta
An agentic-AI perspective on Computer Use 2.0, covering orchestration patterns, tool use, and how browser agents fits production agent stacks.
10 min read
High-Volume Classification with Claude Haiku 4.5: Throughput Numbers
A practical engineering deep dive into Claude Haiku 4.5 classification, covering architecture, tradeoffs, and what production teams need to know about high throughput AI.
8 min read
Using Claude Opus 4.7 to Synthesize Multi-Year Research Archives
A practical engineering deep dive into Claude Opus 4.7 research, covering architecture, tradeoffs, and what production teams need to know about literature review.
6 min read
Streaming Tokens and Graph Events in LangGraph 1.0
LangGraph's streaming primitives let you push tokens, node events, and tool calls to the UI in real time. The four streaming modes explained with examples.
9 min read
LlamaIndex Agentic Workflows: Event-Driven Agents Done Right
LlamaIndex's event-driven Workflows are the cleanest async agent primitive of 2026. A walkthrough with real RAG-with-tools examples and the gotchas to avoid.
11 min read