This project, CCMP, is a GitHub Issues-centric project management system designed for AI-driven development workflows, specifically with Claude Code. It uses bash scripts and markdown configurations to manage specs, dependencies, and task tracking, aiming to reduce friction caused by external PM tools. The creator claims it has halved shipping time internally and is seeking feedback from others in AI-driven or GitHub-centric workflows.
A starter prompt for Claude Code, what you'll need, and how to reach them.
You are an expert full-stack developer. Your task is to build a modern, robust, and fast command-line interface (CLI) tool for AI-driven project management, inspired by 'ccpm'. The core idea is to treat GitHub Issues as the central database for project artifacts (PRDs, epics, tasks) and use AI for brainstorming and initial decomposition. Use Next.js 16 App Router (for any potential web UI if needed later, but focus on CLI first), React 19, Tailwind v4, and the AI SDK v6 with Gemini for AI interactions. The CLI should be built with a performant language like Rust or Go, with a Node.js wrapper for ease of distribution/CLI experience. Data should be stored in Neon Postgres (for configuration and potential caching) and directly in GitHub Issues via its API. Your first MVP slice is to implement the 'Brainstorm PRD & Create Epic/Tasks' feature. This involves: 1. **CLI Command:** A `ccpm init` command that takes a brief project description as input. 2. **AI Interaction:** Uses the Gemini LLM (via AI SDK) to expand the project description into a more detailed, structured Markdown PRD (Product Requirements Document). 3. **GitHub Integration:** Creates a new GitHub Issue (labeled 'Epic') with the generated PRD in its description. Then, it uses the LLM to decompose the Epic into 3-5 initial sub-tasks, creating separate GitHub Issues for each, linking them to the Epic. 4. **Local Configuration:** Store GitHub repo details (owner/name) and API token locally (e.g., in a `.ccpmrc` file or a simple Neon DB table for a more robust setup if a web UI component is introduced later). Focus on robust error handling, clear CLI output, and secure handling of API keys. Provide the full project structure, initial code for the CLI command, the AI prompt for PRD generation, and GitHub API interaction logic. The output should be a fully functional, minimal CLI that can execute this first feature.
Reach solo founders and indie operators interested in automating their build/ship loop and AI development workflows with a more integrated, AI-centric project management tool.
I built a lightweight project management workflow to keep AI-driven development organized. The problem was that context kept disappearing between tasks. With multiple Claude agents running in parallel, I’d lose track of specs, dependencies, and history. External PM tools didn’t help because syncing them with repos always created friction. The solution was to treat GitHub Issues as the database. The "system" is ~50 bash scripts and markdown configs that: - Brainstorm with you to create a markdown PRD, spins up an epic, and decomposes it into tasks and syncs them with GitHub issues - Track progress across parallel streams - Keep everything traceable back to the original spec - Run fast from the CLI (commands finish in seconds) We’ve been using it internally for a few months and it’s cut our shipping time roughly in half. Repo: https://github.com/automazeio/ccpm It’s still early and rough around the edges, but has worked well for us. I’d love feedback from others experimenting with GitHub-centric project management or AI-driven workflows.
Reply in the Hacker News thread to aroussi, then email them if a contact address is provided in their profile or repo.
“I've been building out a robust, modern CLI version of a GitHub Issues-centric PM tool, inspired by your `ccpm`. I've already prototyped the AI-driven PRD and task decomposition using Gemini, and it's fast and integrated. Would you be open to a quick demo and sharing some feedback?”
Open the original ↗