Evals for Claude Agents: Measuring Quality & Gating Ship
Build an eval loop for agentic Claude systems: task-level checks, LLM-as-judge, trajectory metrics, and CI gates that block regressions before they ship.
7 min read
Agentic AI & LLMs
Agentic AI & LLMs archive page 169 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
Securing Claude Agents: Sandboxing & Injection Defense
Harden Claude agents against prompt injection: sandbox tool execution, enforce least privilege, keep secrets host-side, and bound the blast radius.
7 min read
Securing Claude Agents: Sandboxing & Least Privilege
Harden agentic Claude systems with sandboxing, least-privilege scoping, secret handling, and prompt-injection defense. Break the lethal trifecta.
7 min read
Claude Agent Token Costs: Caching, Batching, Speed
Cut Claude agent costs with prompt caching, the Batches API, effort tuning, and model routing — keep runs cheap and fast without losing quality.
7 min read
Cutting Claude Agent Token Costs: Caching & Batching
Keep agentic Claude runs cheap and fast with prompt caching, per-step model routing, batching, and context trimming — without losing quality.
7 min read
Debugging Claude Agents: Loops, Bad Tool Calls, Fixes (Claude Api Skill Ecosystem)
Diagnose and fix the failure modes Claude agents hit in production: runaway loops, wrong tool calls, and hallucinated arguments — with concrete fixes.
8 min read
Debugging Claude Agents: Loops, Bad Tool Calls, Fixes (Product Development Agentic Era)
Fix the three failure modes that break Claude agents in production: runaway loops, wrong tool calls, and hallucinated arguments. A practical debugging guide.
8 min read
Claude Prompt & Context Design: What to Keep, What to Cut
Design context for Claude agents: what to keep, what to load on demand, what to leave out, and why cutting content sharpens reasoning and cuts cost.
7 min read
Prompt and Context Design for Claude Agents That Last
Context engineering for Claude agents: the least-context principle, what earns a place in the window, what to evict, and compaction for long tasks.
6 min read