Spec Driven Development for Developers and AI Agents.
Turn endless iterations into one-shot features.
Shotgun writes codebase-aware specs for AI coding agents. We index your entire repo, research how features should fit your architecture, and generate technical specs that keep Cursor, Claude Code, and Lovable on track.
Export to AGENTS.md. Your AI agent builds correctly the first time instead of derailing halfway through.
Open source CLI for developers working with AI coding agents.
No Cap
Love this!
Shotgun CLI
@ednevsky Thanks Alexander! 🙌 Glad it resonates
Shotgun CLI
@ednevsky We’ve just released the beta version (codebase understanding, export to agents.md, modes - research, spec, planning, tasks, export).
Up for testing? https://github.com/shotgun-sh/shotgun-alpha
Zencal
Let's go 🚀 🇵🇱
Shotgun CLI
@octocode 🇵🇱 thank you!
Shotgun CLI
@octocode 🔥
Congrats! 💪 Fingers crossed!
Shotgun CLI
@mateusz_kowalczyk2 thank you so much!!
This looks awesome. I’m wondering how Shotgun fits into an actual dev workflow — do teams typically keep the specs it generates in the repo (like docs), or are they more ephemeral prompts for Cursor/Claude?
Shotgun CLI
@bessie_west spec becomes a licing doc in your repo. We’ve just released the beta version. Up for testing?https://github.com/shotgun-sh/shotgun-alpha
I'm getting started with testing. My first impression is very positive. You don't even realize how many times I got stuck with Loveable or Replit because they generated my assumptions in a roundabout way.
Shotgun CLI
@marszalkowski_ Thanks so much for giving it a spin, Piotr 🙌 That’s exactly the pain we’ve been trying to solve — when agents generate something “almost right” but it drifts from your intent.
The goal with repo-aware specs is to cut down on those roundabout assumptions and make the path from idea → build much straighter.
Would love to hear as you keep testing: which type of artifact (PRDs, trade-offs, architectures) feels most useful in your workflow?
@kamila_dabrowska Thanks for asking! In my workflow, especially when building platforms integrated with databases and leveraging LLMs for processing diverse data types like video, graphics, code and text, I find architectural artifacts and trade-off analyses to be the most useful.
Amazing product, as non technical vibe coder it makes my life much easier!
Shotgun CLI
@mateusz_pawelczuk1 Thanks Mateusz 🤠 Love hearing that — making vibe-coding less painful for non-technical builders was a big part of why we started Shotgun.
Shotgun CLI
@mateusz_pawelczuk1 We’ve just released the beta version (codebase understanding, export to agents.md, modes - research, spec, planning, tasks, export).
Up for testing? https://github.com/shotgun-sh/shotgun-alpha
It's a problem we've come across often and this helps a long way to solve it 👍
Shotgun CLI
@michal_kaczor We’ve just released the beta version (codebase understanding, export to agents.md, modes - research, spec, planning, tasks, export).
Up for testing? https://github.com/shotgun-sh/shotgun-alpha