Learn · Deep Dives
Deep Dives guides
Longer reads on how AI agents behave in real work.
AI Agents in Production Examples: 6 Jobs They Do TodayAI agents in production examples: six jobs agents really do today, from support tickets to on-call, plus a free skill on our shelf to start each one.→When Not to Use AI Agents: 5 Checks Before You ShipWhen not to use AI agents at work: five checks that tell you an agent should stay a prototype, plus 5 free skills that run each check before launch.→How to Reduce LLM Costs Without Hurting QualityHow to reduce LLM costs without worse or slower answers: audit spend, trim context, prove a cheaper setup holds up, and five free skills for the plan.→Is Claude Getting Worse? How to Measure It YourselfIs Claude getting worse, or is it your session? A plain way to measure a quiet model change on your own tasks, plus five free skills to run the checks.→AI Agent Decision Making: When a Model Like JEV HelpsAI agent decision making, answered plainly: when a typed decision model like JEV earns a place in your agent loop, when it does not, and five free skills to check.→AI Agent Use Cases for Personal Use: 5 First Jobs to Give ItAI agent use cases for personal use, picked by first job: inbox, morning brief, meeting notes, research, topic alerts. Five free skills you can open and use now.→Multi-Agent Debate in LLMs: Personas That Reason DifferentlyMulti-agent debate in LLMs only helps when each persona reasons differently. Five steps that make it happen, plus five free skills you can open and use now.→Claude Code subagents explained: what they are, with 5 real examplesWhat Claude Code subagents are, where the files live, and five free, open-source subagent files from real repos, with the limits each file admits.→