Think agentic.https://think-agentic.pages.dev/Field notes on building agentic solutions across the Microsoft stack.en-gbHugo© 2026 Dan FinanSun, 18 Oct 2026 13:30:00 +0100Hello, worldhttps://think-agentic.pages.dev/2026/hello-world/https://think-agentic.pages.dev/2026/hello-world/Sun, 18 Oct 2026 13:30:00 +0100Dan FinanA first post, to check the type, the code blocks and the feed all work. Replace it with something better.

This is a sample post. It shows how each part of a post looks, so you can see the styles before writing anything real. Delete it, or set draft: true, when you publish your first proper post.

A heading

Body text is set in EB Garamond. Paragraphs stay under about 70 characters a line so long posts are comfortable to read. Links look like this.

Quotes are set large and in italic, for the line you want people to remember.

Lists

  • Write in Markdown
  • Save the file in content/posts/
  • Push to GitHub, and the site and RSS feed update

Code

from agents import Agent

agent = Agent(name="Researcher", instructions="Find three sources and summarise them.")
print(agent.run("What changed in Copilot Studio this month?"))

That’s it.

]]>This is a test from the CMShttps://think-agentic.pages.dev/2026/this-is-a-test-from-the-cms/https://think-agentic.pages.dev/2026/this-is-a-test-from-the-cms/Sun, 18 Oct 2026 09:53:00 +0000Dan FinanThis is the summary box.

This is the test post.

Heading 1

  • List
  • List 2
  • List 3

Quote goes here

Code code code...
]]>Agents need boring infrastructurehttps://think-agentic.pages.dev/2026/agents-need-boring-infrastructure/https://think-agentic.pages.dev/2026/agents-need-boring-infrastructure/Sun, 11 Oct 2026 08:00:00 +0000Dan FinanThe demo is the easy part. What makes an agent useful in production is logging, permissions, retries and someone who owns it on a Tuesday afternoon.

Every agent demo looks the same. A prompt goes in, a tool gets called, a tidy answer comes out, and the room nods. Then someone asks what happens when the API times out, and the room goes quiet.

The gap between a good demo and a useful agent is mostly infrastructure. None of it is exciting, and all of it decides whether the thing survives its first month.

What production actually asks for

  1. Identity. Whose permissions does the agent act with? “The service account” is an answer, but rarely a good one.
  2. Logging you can read. Not just that a tool was called, but why, with what inputs, and what came back.
  3. Retries and timeouts. Downstream systems fail. The agent should fail politely too.
  4. An owner. A named person who gets the alert and can switch it off.

If nobody can explain what the agent did yesterday, it isn’t ready to do anything tomorrow.

Start smaller than you think

The most reliable agents I’ve seen do one narrow job, with clear inputs and a human checking the output for the first few weeks. Scope can grow once trust exists. It rarely works the other way round.

The boring work isn’t a tax on the interesting work. It’s what lets the interesting work keep running.

]]>