PF

Pamela Fox

Things Pamela Says on Podcasts

Where to Find Them

Pamela Fox has been a guest on The InfoQ Podcast (2 times) , Test & Code , Hanselminutes with Scott Hanselman and Talk Python To Me .

Recently: “#569: Connecting Agents to Data via MCP, CLI, and More” on Talk Python To Me (November 2026); “AI and visualizing multidimensional vectors with Pamela Fox” on Hanselminutes with Scott Hanselman (June 2025); “Using ChatGPT to Search Enterprise Data with Pamela Fox” on The InfoQ Podcast (March 2024); “Using ChatGPT to Search Enterprise Data with Pamela Fox” on The InfoQ Podcast (November 2023); “199: Is Azure Right for a Side Project? - Pamela Fox” on Test & Code (May 2023).

What They Said

“The results from that particular experiment were that the GitHub CLI was the best. It required fewer tokens overall and was successful more of the time. Part of that is because the LLMs know the GitHub CLI. The GitHub CLI is deep, deep in the weights. They really don't know the GitHub MCP server much at all.” — Pamela Fox, Talk Python To Me

Michael Kennedy asked for her view on the MCP versus command-line debate. She describes a colleague's test where an agent fixed a real bug using the GitHub CLI, the GitHub MCP server or neither. Her advice is to offer all of them and test each with agents.

Talk Python To Me · 2026-11-16 Permalink → Listen →
Talk Python To Me Around 44:44 into the episode
Pamela Fox

Yes, we have to have the required MCP versus CLI debate. We do. It wouldn't be

Speaker 4

complete if we didn't do it.

Pamela Fox

I'm actually obliged. So yeah, it's very interesting, actually. Like my colleague has a project that's now called Scope. And in this project, they actually will set up scenarios to see how well an agent succeeds at a particular task in a particular environment. And the example that he gave in a talk the other day was using the GitHub CLI GH command and the GitHub MCP server and then just nothing at all compared to see like how long would it take? How successful would it be? Like would it, so in this case, it was actually fixing a bug on one of my repos, which I didn't know he was going to do. So I was surprised to see this PR the morning of his talk. But yeah, so he's, you know, so it's comparing to see, you know, token efficiency success, number of tool calls, all that stuff, right? Now, the results from that particular experiment was that the GitHub CLI was the best, right? And I think a lot of people would actually predict that. Like it was, you know, it was, you required less tokens overall and was successful more of the time. But why is that? Part of that is because the agents, like…

Michael Kennedy

My agent runs rough a lot. I don't know about yours, but I told the standing rule for all of my agents, if they're doing coding, that is, is whenever you're done, you're not done until rough format, rough check, and TY check. All of those run without warnings, without errors. You know what I mean?…

Pamela Fox

Right. So that, I mean, you can put that in your agents.md. I oftentimes have that in my agents.md. You can even formalize like a lot of these coding agents have what's called like a hook, at least VS Code, we call it a hook. So you can have a hook that runs every single time and make sure that it…

Michael Kennedy

If you want to be super old school, I mean, we're talking like 2024 old school, you could put it in a Git pre-commit hook, really. Oh, yeah, of course.

Speaker names come from our own diarization. The position is estimated from where the line sits in the episode.

Collections They Appear In