Vibe code like a PRO
Horacio shares an AI dev task tool discovered from a GitHub repo and video, envisioning a workflow where background agents autonomously resolve Linear issues and create pull requests.
found this repo from this video: https://github.com/snarktank/ai-dev-tasks
Can't wait to start playing with this. I'm curious if this can be integrated with Linear 🤔?
What I want
In Seed Hypermedia we do use Linear to track issues, and based on this tooling what I would love is:
- get all the issues assigned to me from Linear
- expand the issue and explain how to fix/solve it based on the codebase
- spawn background agents for each task to solve those
- create PRs when the issue is fixed.
With this setup in place, I can focus more on bigger and more complex tasks and improve the application a lot more.
What do you think? Is there something else you would do to improve this workflow?
---
See also:
- How I Built a Full Booking System as a Side Project — With Devin as My Engineering Partner — the real-world validation of this exact vision: I built a full booking system using Devin as an async engineering partner, with the same fetch→research→PR pipeline from my phone
- How to Background Agents — using Cursor background agents to autonomously solve tasks spawned from Linear issues
- State machines to visualize the work of agents — state machines make spawned agent behavior deterministic, observable, and predictable, directly applicable to the background agent workflow described here
- Product Backlog — Opportunity 26: Linear-to-Agent Pipeline CLI —
@shm/linear-agentCLI packaging the exact 4-step pipeline described here: fetch→expand→spawn→PR - I want to automatically download Bugs from Linear and add them to Seed using Cursor and Python — Seed community support guide showing how to connect Linear → Cursor via MCP today, with step-by-step setup,
mcp.jsonconfig, and automatic PR generation — a working template for the workflow envisioned here - Graduated Trust: A Web-of-Trust Proposal for Open Source in the AI Era — the same AI capabilities enabling efficient background agents also create the "slop PR" problem; graduated trust proposes a social-graph-based solution to separate signal from noise in AI-augmented contribution workflows