Log inSign up
Waldek Mastykarz
35.1K posts
Waldek Mastykarz profile banner
@waldekm

Waldek Mastykarz

@waldekm
Evaluating AI Coding Agents @Microsoft, Agent Experience, #DevProxy, #CLIMicrosoft365, @m365pnp member
Alphen aan den Rijn, NL
blog.mastykarz.nl
Joined December 2008
499
Following
11.3K
Followers
RepliesRepliesRepostsRepostsMediaMediaArticlesArticles

Log in or sign up for X

See what’s happening and join the conversation

Continue with phone
or
Log in with username or email
Terms·Privacy·Cookies·Accessibility·Ads Info·© 2026 X Corp.
  • @waldekm
    Waldek Mastykarz
    @waldekm
    1h
    Here’s to the one-shot prompters. The vibers. The shippers. The prompt-and-prayers. The square pegs who never checked whether there was a round hole. The ones who see a blank chat box and think: production. They’re not fond of iterations. And they have no respect for evals.
  • @waldekm
    Waldek Mastykarz
    @waldekm
    2h
    For all my fellow one-shot prompters
    @waldekm
    Waldek Mastykarz
    @waldekm
    21h
    One Shot His palms are sweaty, tokens weak, context heavy There’s JSON in his output already, schema spaghetti He’s nervous, but on the surface he looks calm and ready To drop prompts—but the model keeps forgetting What he wrote down, the whole context goes so loud He opens his
  • @waldekm
    Waldek Mastykarz
    @waldekm
    2h
    Great read on AI evals! Whether you’re building AI-powered apps, agent extensions like skills or MCP servers or even just using AI coding agents for your work, you should have evals. How else will you tell if the outcome meets your needs?
    @ttorres
    Teresa Torres
    @ttorres
    Sep 3
    AI evals have been the "it" skill for product teams for over a year. I've even called evals a new discovery habit. But I still meet product teams who only have a vague idea of what evals are. And it's not their fault. Most of the writing on this topic is intended for engineers
    4
  • @waldekm
    Waldek Mastykarz
    @waldekm
    14h
    Coding agents can take us from intent to production in minutes. But what happens to the understanding we used to build along the way?
    Image
    What if coding agents had a learning mode?
    From blog.mastykarz.nl
  • @waldekm
    Waldek Mastykarz
    @waldekm
    15h
    Love the annotations! Finally getting a red pen to leave contextual notes for the agent. Awesome work folks!
    @GitHubNext
    GitHub Next
    GitHub
    @GitHubNext
    Sep 2
    A new project from GitHub Next! Chopin is a real-time, multiplayer agentic planning environment. Designed to get your whole team aligned on the plan before you hand it off to agents. We find our current agent planning tools feel cumbersome, static, lonely, and drown us in walls
    1
Advertisement
Advertisement