#coding-agents · AI & ML
Coding agents
Codex, Cursor, Gemini CLI, Aider and other agents that write code for you, with the plugins and instruction files around them.
This week (Nov 1 – Nov 7), the top-ranked are dsh-superagent-prototype (+18,169), skills (+2,194) and dsh-Go (+10,461). The 436 ranked repos gained 144,888 stars, 19% more than the week before.
Updated Nov 7, 15:36 UTC · ranked every hour from GitHub’s daily star counts
Updated 56 min ago
Emerging this week
Listed by acceleration against their usual week; the score is the rankings’ own: stars gained in 7 days ÷ √(stars before, minimum 100).
- 1feimacode/open-design-agent-kit Make posters, social posts, slides and landing pages in the coding agent or chat you already use (Copilot, Claude, Codex, Cursor), with 198 design systems, checks and PDF/PNG/PowerPoint export. Breakout week (~1 a week)Gained in 7 d +401 Score 15.9 Total 1k
- 2Muvon/octomind Open-source AI coding agent and agent runtime: one binary, any model, MCP-native. Runs in terminal, CI, or as a daemon. 50× its usual pace (~8 a week)Gained in 7 d +398 Score 28.3 Total 596
- 3YizhengWw/HachimoDock A desktop pet that reacts to your Claude Code sessions in real-time — thinking, typing, juggling, sleeping, and more. 75× its usual pace (~5 a week)Gained in 7 d +377 Score 16.1 Total 928
- 4guregodevo/memdoor AI coding agent that makes a team of small models beat a frontier one Breakout week (~1 a week)Gained in 7 d +334 Score 33.4 Total 374
- 5 Gained in 7 d +323 Score 22.2 Total 535
- 6solongate/agent-security Local-first, open-source security gateway for AI coding agents. Enforce tool-call policies, detect secrets, and audit actions on-device. Breakout week (~2 a week)Gained in 7 d +283 Score 24.0 Total 422
- 7MicTx/spec-harness 面向 AI 编程智能体的任务包工作流工具,用可验证的范围、证据和 Git 门禁管理长期改动。(Task-package workflow for AI coding agents with verifiable scope, evidence, and Git gates.) 89× its usual pace (~3 a week)Gained in 7 d +266 Score 26.6 Total 277
- 8FinanceFlash/unvibecode Complex code hides connections, workflows, and business risks.Unvibe complex code. Trace business workflows. 19× its usual pace (~24 a week)Gained in 7 d +467 Score 16.2 Total 1.3k
- 9uniquechao/PandaDOS Your self-hosted control plane for Claude Code and Codex. Turn issues into queued, observable, approval-aware coding workflows—from any browser. After quiet weeksGained in 7 d +194 Score 14.9 Total 363
- 10wanshuiyin/HERO-Anti-OverDefense HERO = Hashing · Edge cases · Rubrics · Overbuild — the four shapes coding agents over-defend in. A paste-in contract that stops them. Works with Claude Code, Codex, Antigravity, Cursor, Copilot, Windsurf, Gemini CLI. 19× its usual pace (~14 a week)Gained in 7 d +265 Score 11.5 Total 800
Back to the ranking Past weeks: Oct 26Oct 19
About this topic
Some of these agents work inside an editor, others in a terminal, OpenCode, Cline, Roo Code and GitHub Copilot among them; the projects built around them also include harnesses that drive an agent. Claude Code and its add-ons have their own topic. AI agents covers agents for any task, while this one keeps to those that program. For developers comparing agents, or after instructions and add-ons for the one they already use.