Automation and Agents
DeepSeek V4.1 Flash: Cost per Successful Agent Task
DeepSeek V4.1 Flash's low rates and smaller reported cache make it worth testing, but your task cost depends on quality, retries, tools, and review.
Read the articleNotes from the work
Useful ideas, with the messy parts left in.
Build notes, practical systems, and a closer look at what changed. Written to help you make a decision, try something useful, or avoid learning the annoying part twice.
Showing 51 to 60 of 130 articles
Keep reading
More from the notebook.
Automation and Agents
OpenAI Agents API Sandbox: Hosted, Self-Hosted, or No Environment?
The Agents API separates its managed harness from the place code and tools run. Choose no environment, OpenAI-hosted, or self-hosted by deciding what your task needs and what your team can safely operate.
Read the article
Automation and Agents
OpenAI Agents API vs. Responses API: Which Should You Use?
OpenAI now offers a managed Codex harness through Agents API. Here is how to decide whether your workflow needs it, an application-owned SDK loop, or the direct Responses API.
Read the article
AI in Practice
Supervise a Claude Managed Agents Session from the Terminal
A practical guide to connecting, observing, intervening in and verifying a hosted Claude Managed Agents session.
Read the article
AI in Practice
Claude Managed Agents Auto Permissions: How to Pilot Them Safely
Understand what Claude Managed Agents auto permissions decide and how to pilot them without mistaking evaluation for human approval.
Read the article
Automation and Agents
Can You Trust a ChatGPT Work Data Agent Dashboard?
A practical review framework for checking Data agent evidence, calculations, connected-account access, published dashboard contents, audience, and downstream actions.
Read the article
AI in Practice
How to Plan a ChatGPT Work Data Agent Pilot
A practical pilot brief for choosing a decision, source, metric, time period, and trusted comparison before asking ChatGPT Work Data to analyze company data.
Read the article
Systems and Workflows
Claude Code Projects: A Review Plan for Parallel AI Work
A practical ownership map for the decisions that remain human when Claude Code Projects coordinates multiple threads.
Read the article
Automation and Agents
Claude Code Projects: When Parallel AI Work Is Worth It
A practical test for deciding whether Claude Code Projects fits a real multi-step software effort, plus a scorecard that separates coordination gains from extra review and usage.
Read the article
Automation and Agents
How to Use Siri AI's Personal Context With Clear Boundaries
Cross-app context can make a personal assistant more useful, but it also raises a practical question: what information does this task need, and where should a person take over?
Read the article
Follow the work
A useful note when there is something worth sending.
New builds, honest lessons, and practical ways to make more room for the work that needs you.
Free to read. Unsubscribe at any time. Privacy