zclllyybb

zclllyybb / multi-agent-todo

Public

A persistent, daemon-based multi-agent system that leverages opencode to automatically discover, plan, implement, and review code changes across a codebase.

agentic-ai agentic-workflow ai ai-agent framework
15
0
100% credibility
Found Mar 03, 2026 at 15 stars -- GitGems finds repos before they trend. Get early access to the next one.
Sign Up Free
AI Analysis
Python
AI Summary

An AI-powered system that scans codebases for TODO comments, prioritizes them, and uses specialized agents to plan, implement, review, and deliver fixes via a web dashboard.

How It Works

1
🔍 Discover the tool

You find this smart helper that automatically finds and fixes TODO notes in your code project.

2
📁 Point to your code

You tell it where your code folder is and pick friendly AI thinkers to do the work.

3
🔎 Scan for TODOs

It searches your project and lists all the pending tasks with smart scores on ease and effort.

4
⭐ Review the backlog

You see a clear prioritized list of fixes, with notes explaining why each is ready or tricky.

5
🚀 Start the fixers

You pick tasks and let the planner, coder, and reviewers work together in parallel.

6
📊 Watch progress live

The dashboard shows real-time updates as they plan, code, check, and retry until perfect.

✅ Get approved fixes

You receive clean code branches ready to merge, with your TODOs turned into reality.

Sign up to see the full architecture

5 more

Sign Up Free

Star Growth

See how this repo grew from 15 to 15 stars Sign Up Free
Repurpose This Repo

Repurpose is a Pro feature

Generate ready-to-use prompts for X threads, LinkedIn posts, blog posts, YouTube scripts, and more -- with full repo context baked in.

Unlock Repurpose
AI-Generated Review

What is multi-agent-todo?

This Python daemon runs persistently in the background, leveraging opencode to automatically discover TODO and FIXME comments across your entire codebase, score them by feasibility and difficulty, then plan, implement, and review fixes in isolated git worktrees. Tasks process in parallel through a planner-coder-reviewer pipeline with retries and multi-model support, all managed via CLI commands like `python cli.py scan` or a real-time web dashboard for task dispatch, revision, and publishing branches.

Why is it gaining traction?

Unlike basic codegen tools, it handles the full cycle from discovery to review-only mode for PRs or patches, with human-in-the-loop feedback that loops back into retries—preventing half-baked changes. Parallel execution across worktrees avoids conflicts in large repos, and runtime model swaps (e.g., cheap models for simple tasks) keep costs down while ensuring all reviewers approve before completion. The dark-themed dashboard makes monitoring agent runs and git status effortless.

Who should use this?

Backend maintainers drowning in legacy TODOs across sprawling codebases, open-source project leads automating low-hanging fruit fixes, or teams experimenting with AI agents for routine refactors without disrupting main branches. Ideal for Python, C++, or Java repos where opencode is already set up.

Verdict

Promising for daemon-based multi-agent automation on code changes, but at 15 stars and 1.0% credibility, it's early-stage—docs are solid via README and screenshots, but expect tweaks for production. Try it on a side repo if you want persistent TODO hunting; skip for mission-critical work until more battle-tested.

(198 words)

Sign up to read the full AI review Sign Up Free

Similar repos coming soon.