HackerReposA FIELD GUIDE TO OPEN SOURCE
ProSponsor the shelf♧ 0
← Back to top all-time
earendil-works
pi
TypeScript2025–2026
earendil-works / pi

pi

AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

82Distinct people
101Mentions
2Explicit recommenders
121.7KGitHub stars
REPOSITORY AT A GLANCE

What to know before exploring

What it is

AI agent toolkit: unified LLM API, agent loop, TUI, coding agent CLI

GitHub

Where it is used

  • OpenClaw — OpenClaw embeds the Pi SDK to run an AI coding agent inside its messaging Gateway. It uses Pi’s createAgentSession-based runtime to control sessions, tools, and streaming without external processes. Source

What people are saying

I love Pi . I have been building an agent using its agent harness https://github.com/earendil-works/pi/tree/main/packages/agen... . I love the simplicity .

manoji used a recommendation phrase. 2 indexed comments match. This is a phrase match, not a verdict.

No classified warning is attached to this page. Read the discussions before treating silence as approval.

Also in these discussions

Mentioned in the same Hacker News threads. That is not proof they are substitutes.

This week on Hacker News

Nov 9–Nov 16 UTC · Direct repository links and separately labeled related project stories. Indexed through 11/16/2026, 7:27:27 AM UTC.

Pi: Virtual Models direct repository link

Read the discussion on HN. A full-thread summary is not available yet.

Mentions over time

HN mentions by month101 dated mentions

Observed mentions: 2025-11-16 20:54:39 UTC to 2026-10-29 13:42:01 UTC.

01020302025-11-01 00:00:00 UTC to 2025-12-01 00:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 commentsNov 20252025-12-01 00:00:00 UTC to 2026-01-01 00:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 comments2026-01-01 00:00:00 UTC to 2026-02-01 00:00:00 UTC (exclusive): 5 mentions; 4 submissions, 1 comments2026-02-01 00:00:00 UTC to 2026-03-01 00:00:00 UTC (exclusive): 28 mentions; 4 submissions, 24 comments2026-03-01 00:00:00 UTC to 2026-04-01 00:00:00 UTC (exclusive): 8 mentions; 2 submissions, 6 comments2026-04-01 00:00:00 UTC to 2026-05-01 00:00:00 UTC (exclusive): 10 mentions; 0 submissions, 10 commentsApr 20262026-05-01 00:00:00 UTC to 2026-06-01 00:00:00 UTC (exclusive): 6 mentions; 0 submissions, 6 comments2026-06-01 00:00:00 UTC to 2026-07-01 00:00:00 UTC (exclusive): 1 mentions; 0 submissions, 1 comments2026-07-01 00:00:00 UTC to 2026-08-01 00:00:00 UTC (exclusive): 8 mentions; 0 submissions, 8 comments2026-08-01 00:00:00 UTC to 2026-09-01 00:00:00 UTC (exclusive): 13 mentions; 4 submissions, 9 comments2026-09-01 00:00:00 UTC to 2026-10-01 00:00:00 UTC (exclusive): 4 mentions; 0 submissions, 4 comments2026-10-01 00:00:00 UTC to 2026-11-01 00:00:00 UTC (exclusive): 16 mentions; 2 submissions, 14 commentsOct 2026

Counts come from the repository’s dated source evidence. Index cutoff: 2026-11-02 08:16:59 UTC.

View mention counts and date ranges
UTC ranges include their start and exclude their end. Select a period to filter the source evidence.
PeriodStart (UTC)End (exclusive, UTC)MentionsSubmissionsComments
Nov 20252025-11-01 00:00:00 UTC2025-12-01 00:00:00 UTC101
Dec 20252025-12-01 00:00:00 UTC2026-01-01 00:00:00 UTC101
Jan 20262026-01-01 00:00:00 UTC2026-02-01 00:00:00 UTC541
Feb 20262026-02-01 00:00:00 UTC2026-03-01 00:00:00 UTC28424
Mar 20262026-03-01 00:00:00 UTC2026-04-01 00:00:00 UTC826
Apr 20262026-04-01 00:00:00 UTC2026-05-01 00:00:00 UTC10010
May 20262026-05-01 00:00:00 UTC2026-06-01 00:00:00 UTC606
Jun 20262026-06-01 00:00:00 UTC2026-07-01 00:00:00 UTC101
Jul 20262026-07-01 00:00:00 UTC2026-08-01 00:00:00 UTC808
Aug 20262026-08-01 00:00:00 UTC2026-09-01 00:00:00 UTC1349
Sep 20262026-09-01 00:00:00 UTC2026-10-01 00:00:00 UTC404
Oct 20262026-10-01 00:00:00 UTC2026-11-01 00:00:00 UTC16214

What Hacker News says

13 sources
alexvboe·Aug 30, 2026Submission

Show HN: Cogram Studio – CAD and BIM workspace for humans and agents

Hi HN, Rick and Alex here, co-founders of Cogram. We’ve been making project-management software for architects and engineers since 2023, and are now experimenting with a second product. Cogram Studio (studio.cogram.com) is a CAD and BIM workspace for AI agents to create three-dimensional models and dimensioned drawings. Studio runs FreeCAD (https://github.com/FreeCAD/FreeCAD) 1.1 headlessly, using the OpenCASCADE (https://github.com/Open-Cascade-SAS/OCCT) geometry kernel. It imports and exports STEP, IFC, STL, DXF and FCStd files. Studio exposes an MCP server, so you can operate it from Claude Code, Codex, ChatGPT, or other agents. Tools cover scripting against FreeCAD, inspecting the model visually, CRUD operations on views, tables, and sheets, importing terrain, etc. Studio also provides a built-in agent based on Pi (https://github.com/
Read the rest
earendil-works/pi) and a basic UI for 3D model inspection, measurement, view, table, and sheet creation. There are two ways to try it: 1. Bring your own agent. Point Codex or Claude Code at studio.cogram.com/skill.md and ask it to follow instructions there. This starts a one-hour Studio session and opens the modelling environment in your browser, connected to your agent. You need to confirm sign up only if you want to save the session. 2. Use Studio’s built-in agent. Submit a prompt at studio.cogram.com and sign up. You receive 50 free credits. Here are two models created by agents: - A simple bookshelf (https://studio.cogram.com/view/shr_1agb21nd4) with two sheets - The Colosseum (https://studio.cogram.com/view/shr_3tm32c64y) [large, can take minutes to load], with two sheets showing dimensioned and annotated views. For now, Studio works best when you use it iteratively. If you ask an agent for a finished, complex model and “let it rip”, it may produce something that looks convincing but doesn't hold up to scrutiny. Agentic modelling feels roughly like where agentic programming was in 2024: useful for speeding up repetitive or bounded tasks but not yet reliable enough to complete long, complex jobs. Recent models have improved at CAD. We expect progress to accelerate as models become even better at reasoning across the different representations used in architecture and engineering: geometry, drawings, images, materials, physical constraints, specifications, regulations, code, etc. Early users and our own experience suggest that agents are already useful design collaborators. They can help with repetitive CAD work, fabrication options, part sourcing and cost estimates. We expect that as models improve, agents will be able to explore many different designs, run simulations, and check constraints across an entire project. CAD will become a shared workspace for humans and agents. Because design decisions made in CAD shape much of what is eventually built, even modest improvements could have wide effects. We’d love to hear your feedback and ideas if you give Studio a try. We’re also hiring across product, engineering and commercial roles. If you’re interested in working on interesting problems involving AI and the physical world, please take a look at our open positions at https://cogram.com/about#careers.
6 points · Open discussion

Show HN: Always responsive UI for a coding agent

In my limited personal experience, I find the UI of coding agents, or rather their clients, unsatisfying for the following reasons: - They tend to have poor performance characteristics - They make it hard to actually read or copy from the session history - They make it hard to paste large contents, and to undo that paste. - They lack a sane default for sandboxing It appears to me that those UI try to re-use patterns from chat UIs, but fail to address the key difference: large and constantly changing content. For the last couple of months I've been using pi.dev, and while I appreciate the minimalist design of the harness, the default TUI on Mac OS is frustrating: it hogs the CPU as the session gets large, history keeps shifting under your nose making reading or copying from it hard, and pasting into the prompt input is unresponsive. Those
Read the rest
problems compound: I once tried to copy a single line from the session history, while being scrolled away from the bottom, at which point the content on screen shifted, and then when I pasted in the input, the whole thing just took forever without giving intermediary feedback, because the copy had been applied not to my selection but to some large chunk of history (due to the content shift). Then the sandboxing: if you search for solutions, you find things that either are limited, like only sandboxing the bash tool use (while pi is meant to be extended with arbitrary code), or seem more complicated than is necessary by using container software. So on a Friday, I filed https://github.com/earendil-works/pi/issues/7730, and then over the weekend ended-up writing my own Mac OS client to headless pi, with built-in sandboxing. It's been a week now, and it is what I am using, and improving, on a daily basis: https://github.com/gterzian/uni03C0 Some highlights: - In the main rendering path: only process that part of content (whether it is session history or pasted content) that you really need. - Never block the UI. - Keep the UI stable, and leave it entirely idle when not following the streaming content at the bottom. - Everything can be aborted, including a large incoming paste. - Mac OS native sandboxing applied by default (unfortunately using a deprecated API, but one that for now is still what major browsers rely on); configurable by the end user. So this is a fast, resource efficient, always responsive, and sandboxed by default (without needing a container), Mac OS client to headless pi. Please let me know what you think.
3 points · Open discussion
> Using harnesses that are “less chatty” (more token efficient), or tuning existing harnesses to generate less token overhead. agree, this works, undervalued! look at minimal agents that protect the context window: - pi (https://github.com/earendil-works/pi) - smol (https://github.com/smol-env/smol) some thoughts on the other tips (for coding): 1) stronger models are more token efficient for open ended tasks because at the limit … - stronger models can solve tasks that the weaker models can not solve - stronger models make fewer mistakes, compose things better (cli, abstractions, …) - navigate the code base better - are better at removing and simplifying the code base again that of course is difficult to benchmark, so most attention goes to simple benchmarks that show cheaper models can get similar results on 'closed' tasks
Read the rest
with easy to 'eval' results 2) dynamic request and task routing sounds great/obvious but is very very hard - to benefit from caching you don't want to switch model or inference endpoint - to _know_ a certain request can be routed to a weaker/cheaper model needs good context and a strong model to get right and often is still unknowable because the active coding session can go many ways and turn from trivial to challenging in a few turns, always in motion is the future, if you get it wrong you are back in the problem space of #1 using cheaper models and auto-routing do work well for 'closed' tasks where you have something repeatable and can evaluate whether a certain quality threshold is reached that you are comfortable with for open ended coding sessions it is not so easy that said: cheaper does not have to mean weaker, you want to look at the pareto frontier and stay up to date on new good models there are many models like deepseek v4 flash and luna that are both cheaper and way better than most other models
In Managing AI Coding Costs at Scale

MAKE YOURSELF AT HOME

A few shortcuts

/
Search the collection
?
Open this guide
Esc
Close dialogs and navigation

Your saved repositories and interests stay in this browser.