Entering Beta: Codebase Understanding + Agents.md Export + Modes
If generating code is a commodity, then what is built, and how is the most critical question to answer.
This is the goal of Shotgun - to help you turn technical research and spec generation into context for software engineers, AI code-gen tools like Codex, Cursor, Claude Code, with complete codebase understanding, and agents doing the heavy lifting.
Shotgun produces clean, reusable artifacts and exports to the agents.md ecosystem to help you get the most out of code-gen tools and Agents.
New release highlights:
🌵: Modes: Research → Specify → Plan → Tasks → Implement
Hit Shift+Tab to summon different expert advisors for each stage. Researcher, spec writer, architect, task planner — all at your service. Switch freely. Shotgun adapts to you, not the other way around.
📦 Export to agents.md
Every artifact (specs, plans, tasks) now exports into the agents.md ecosystem. That means clean, reusable inputs for Cursor, Claude Code, GeminiCLI, GitHub Copilot, and other agentic codegen tools.
🧠Codebase Understanding
Shotgun reads your repo first: architecture, naming, patterns. Then it gives advice that fits your project. Like pairing with someone who spent a weekend studying your code. Less context-switching, more building.
30-sec to try. More to come. Your code-gen agents will thank us later. WE ARE LOOKING FOR FEEDBACK!
Install instructions: https://github.com/shotgun-sh/shotgun-alpha
Replies
GraphBit
Love how Shotgun structures the research → spec → plan stages before code, context is everything. How well does it adapt when the repo is very large or messy, with inconsistent naming and patterns?
Shotgun CLI
@musa_molla great question. With our codebase understanding engine we parse large codebases into a graph database first which our AI Agent knows how to query efficiently. Please feel free to try it and let us know what you think, I'd love to hear any feedback.