Nick Strayer

Positron Notebook Editor – Positron Use the Positron Notebook Editor, currently in public alpha, for a native notebook experience with integrated AI assistance, data exploration, and improved data science workflows.

Need to dig in further? Just click the "Open in Data Explorer" button in the top right!

Try it out in the latest Positron daily build and let us know what you think!
positron.posit.co/positron-not...

New feature in the Positron Notebook Editor: Inline data-explorer view. Whenever you print a pandas or polars dataframe you get an inline interactive view powered by Positron's data explorer. This allows you to view and sort massive dataframes.

The interesting part about this though is that this is entirely a client-side app/PWA. You add your anthropic key and it stores it to localstorage and then directly makes requests to the anthropic api. No server needed. Hosted on github pages. BYOK

nickstrayer.me/aisle-list/

It takes a photo of a grocery list and uses Claude to turn it into a check list arranged by the optimal path through our typical grocery store (kroger.)

My wife hates apps in favor of hand-writing grocery lists. I have not the attention span to remember what I have already procured from the list or how to best navigate our grocery store to get the items. I built an app to solve this…

Entirely possible that my workflow though was highly fit to 4.5 and I need to figure out how to work with a new model. If that is the case I really hope we don't have to tweak workflows every 3 months as a model comes out. It's exhausting.

After using Opus 4.6 for a while: I dont necessarily think it's better than 4.5. I've noticed more designs that it makes that are kind of obvious "I would not do it that way, let me fix it" than I had with 4.5. #ai #claude

Announcing the Positron Notebook Editor for Jupyter Notebooks - Posit The Positron Notebook Editor in alpha provides a "batteries-included" experience that combines the simplicity of Jupyter with the power of a purpose-built data science IDE.

I got my start in programming with Jupyter notebooks and so they are dear to my heart. If you’re an adventurous data scientist, download the latest Positron and give it a try! Feedback is welcome and highly encouraged!
posit.co/blog/announcing-the-positron-notebook-editor-for-jupyter-notebooks/

For example, the notebooks have best-in-class AI support by leveraging full notebook context and Positron’s existing kernel/interpreter infrastructure

For more than a year I have been working on a brand new Jupyter Notebook editor for Positron. This is a ground-up build of a new Jupyter Notebook experience built to leverage all the knowledge and tools Posit/Positron brings to the data science table. 🧵#jupyter

GitHub - nstrayer/simple-commands-mcp: Create an mcp server with a simple config to run commands including automatic long-running processes management. Create an mcp server with a simple config to run commands including automatic long-running processes management. - nstrayer/simple-commands-mcp

github.com/nstrayer/sim...

Using it is as simple as creating a config and pointed to it with an npx command: `npx simple-commands-mcp @config.json`

To help with this I built a simple MCP server that gives agents a structured way to manage these processes. Works for any bash command really. Fun experiment in making AI tools more practical for actual development workflows.

I've been working with #AI coding agents lately and noticed they struggle with long-running dev tools like watchers - either they get stuck waiting for a task to finish that never will or they can't see when things fail or need restarting.

Positron is just incredible.
I just tried it:
both R session and Python session in parallel
data explorer with distribution and % of missing at glance
git repo with vizualisation of the branches,
plots and viewer in the bottom right corner.
Just Perfect.
bsky.app/profile/hadl...

PR Notifier menu bar application showing 3 pull requests awaiting review. The dropdown displays two sample requests: '[SAMPLE] Add new dashboard feature' from sample/repo (#123) and '[SAMPLE] Fix login bug on Safari' from another/project (#456). Each PR has 'View on GitHub' and 'Dismiss' buttons. The menu shows 'Last checked: less than a minute ago' and includes tabs for 'Pull Requests (3)' and 'Settings'.

I am terrible about keeping up with PRs I'm supposed to review so I wrote a little menubar app for macs to alert you when you are requested to review a PR. Right in your face so you can't ignore it. github.com/nstrayer/pr-...

I recent got a Fitbit again for a study I’m taking part in. Reminded me of this fun side project from grad school: a year as told by Fitbit. livefreeordichotomize.com/posts/2017-1...
Fun fact: a whole class of Stanford students analyzed my heart rate data for an assignment.

A screen recording of the RStudio IDE. After highlighting some code, a dialog box pops up with a selection of keywords; after selecting one called "cli", some erroring code is rewritten. Then, after highlighting a whole function and selecting a new keyword "roxygen", roxygen documentation is drafted on-the-fly.

I've been working a new #rstats package called pal🤖

Pals bring a library of LLM prompts into your RStudio/Positron sessions, editing code directly at the press of a keyboard shortcut.

Read more: simonpcouch.github.io/pal/

Since I don't want to give the whole world the ability to use my api keys, if you want to play with the app you'll need to run it locally. I tried to make the process as easy as possible though in the README!

Screenshot of a web application with two panes connected by an arrow. The left pane says "R" and contains a basic R script, the right side says "Python" and contains a similar script but written in the python programming language.

I have been slow to get on the AI train but I recently built a demo app to explore the idea of programming language translation using LLMs. The hypothesis being that adding context about the target environment can help. github.com/nstrayer/pol... #ai #rstats #python

Oof spoke before the attributions slide! Work with
@lucystats.bsky.social and @travisgerke.bsky.social !

@malcolmbarrett.malco.io giving a great and super approachable talk on causal inference in R. Awesome work he’s doing with @sherrirose.bsky.social