GitHub Trending AI Developer Tools: 2026-05-14 Ranking and Engineering Signal

Three repositories deserve immediate inspection: obra/superpowers (composable agent workflow skills with over one hundred eighty-nine thousand total stars and the largest community in this ranking), mattpocock/skills (opinionated engineering skills for coding agents, gained over three thousand three hundred stars in one day, MIT license), and github/spec-kit (GitHub's official spec-driven development toolkit with over one thousand one hundred stars gained and support for over thirty AI coding agents). tinyhumansai/openhuman leads the daily ranking with over one thousand six hundred stars gained but carries GPL-3.0 copyleft risk. rohitg00/agentmemory continues its multi-day trending run as the leading persistent memory layer for coding agents. trycua/cua provides open-source sandbox infrastructure for computer-use agents across any operating system.

All star counts and push timestamps reflect GitHub API state as of the enrichment window. The "Stars Gained" column represents the daily delta reported by GitHub Trending; it is a short-window attention metric, not a durability signal (GitHub Trending). Total star counts are provided for enriched repositories only (the top five). Repositories without enrichment data show N/A in the Total Stars and License columns.

Cluster 1: Agent Skills, Specification Tooling, and Workflow Infrastructure

The dominant signal in this cycle is the convergence of agent workflow tooling around structured skill definitions and specification-driven development. obra/superpowers leads with over one hundred eighty-nine thousand total stars and gained over one thousand four hundred stars today. It provides composable skill definitions for AI agent workflows covering brainstorming, dispatching parallel agents, systematic debugging, test-driven development, and writing implementation plans. The philosophy is model-agnostic and agent-agnostic: skills are markdown-based instruction sets that any compatible coding agent can consume. The MIT license and established maintenance history make this a low-risk inspection target for teams building agent-assisted engineering processes.

mattpocock/skills gained the highest star delta among all nineteen trending repositories at over three thousand three hundred stars. The repository focuses on what Matt Pocock calls "real engineering, not vibe coding" — structured skill definitions that prevent agents from losing intent, being verbose, or creating architectural decay. Skills cover engineering workflows like issue triage, domain document layout, TDD loops, and zoom-out architectural reviews, plus productivity patterns like compressed communication and skill authoring. The README explicitly positions these skills against process-heavy alternatives like GSD, BMAD, and Spec-Kit, emphasizing small, composable, and adaptable design. The MIT license and the author's established reputation in the TypeScript community make this a high-value, low-risk reference.

github/spec-kit appeared with over one thousand one hundred stars gained in a single day. This is GitHub's official open-source toolkit for Spec-Driven Development (SDD), a methodology that treats specifications as executable artifacts rather than disposable documentation. The Specify CLI provides a structured workflow: define principles (constitution), write specifications, create technical plans, generate task breakdowns, and execute implementations. It supports over thirty AI coding agents including Claude Code, Cursor, Codex CLI, and Gemini CLI. The README documents an extensive community extension ecosystem with integrations for Jira, Azure DevOps, GitHub Projects, and Linear. The MIT license and GitHub's institutional backing make this a significant new entry in the agent workflow tooling space.

Cluster 2: Personal AI Agent Harnesses and Memory

tinyhumansai/openhuman holds the top trending position with over one thousand six hundred stars gained today and a total exceeding six thousand three hundred. Written in Rust with a Tauri desktop shell, it positions itself as a "Personal AI super intelligence" — a local-first agent that integrates with third-party services via OAuth, runs a Memory Tree knowledge base backed by SQLite, and compresses token usage through a feature called TokenJuice (claiming up to eighty percent reduction). The README includes a feature comparison table against other agent harnesses. The project is in early beta with active daily commits (v0.53.45 as of today). However, the GPL-3.0 copyleft license and ninety-three open issues mean this is inspection territory, not production-ready infrastructure.

rohitg00/agentmemory continues its multi-day trending run with over eight thousand total stars and an Apache-2.0 license. It provides a persistent memory server for coding agents built on the iii engine, offering a four-tier memory consolidation pipeline (Working to Episodic to Semantic to Procedural), triple-stream hybrid search, and over fifty MCP tools. This repository has been covered in a dedicated deep dive on this site; readers should refer to that analysis for architecture, benchmarks, and adoption risks.

Cluster 3: Computer-Use Agent Infrastructure, TTS, and Education

trycua/cua is an open-source infrastructure project for computer-use agents. It provides agent-ready sandboxes for any operating system (Linux, macOS, Windows, Android) via a unified Python API. The project includes CuaBot (co-op computer-use for any agent), Cua-Bench (benchmarks and reinforcement learning environments), and Lume (macOS virtualization on Apple Silicon). The MIT license and cross-platform design make it relevant for developers building GUI-automation agents that need safe, isolated execution environments.

supertone-inc/supertonic gained over eight hundred fifty stars as a Swift-based on-device multilingual text-to-speech engine running via ONNX. While not a core AI developer tool, it represents the growing trend of embedding AI inference directly into applications without cloud dependencies.

rasbt/LLMs-from-scratch continues its multi-day trending presence as a foundational educational resource authored by Sebastian Raschka and published by Manning. It covers coding GPT-like models from the ground up with implementations of Llama, Qwen, Gemma, and Olmo architectures. The developer community values foundational literacy as the agent ecosystem accelerates.

Inspect first: obra/superpowers. With the largest existing community in this ranking at over one hundred eighty-nine thousand stars, this repository represents a battle-tested approach to structuring agent workflows. The skill definitions cover the full software development lifecycle. Even teams that do not adopt the specific skills directly can use them as reference patterns for their own agent instruction sets.

Inspect second: mattpocock/skills. The README articulates agent failure modes with unusual clarity — misalignment, verbosity, broken code, and architectural decay. Each failure mode maps to a specific skill. Even if you do not adopt the specific skill definitions, the diagnostic framework for understanding why agents produce bad output is valuable. The install path is non-destructive via npx skills@latest add mattpocock/skills.

Inspect third: github/spec-kit. A specification tool from GitHub itself, with support for over thirty AI coding agents and a growing community extension ecosystem, is a strong institutional signal. The SDD methodology it promotes (specify, plan, tasks, implement) is becoming a de facto standard in AI-assisted development workflows. Evaluate whether it fits your team's existing process before adopting.

Evaluate cautiously: tinyhumansai/openhuman. The Rust-based agent harness, token compression, and integration breadth are interesting, but GPL-3.0 copyleft, ninety-three open issues, and early beta status mean this is early-adopter territory. Monitor the release cadence and issue resolution rate before investing integration effort.

Worth watching: trycua/cua. Computer-use agent infrastructure is a nascent but growing category. The cross-platform sandbox design and MIT license make it worth monitoring for teams building GUI-automation workflows.

Do not inspect for AI dev tool use: yikart/AiToEarn (AI content marketing platform), CloakHQ/CloakBrowser (anti-detect Chromium with dual-use risk — flagged for mention-only with risk context), influxdata/telegraf (metrics agent), apernet/hysteria (network proxy), ton-blockchain/acton (blockchain toolchain), Greedeks/GTweak (Windows setup tool), ArthurBrussee/brush (3D reconstruction), and imthenachoman/How-To-Secure-A-Linux-Server (Linux security guide). These repositories trended but do not solve AI-developer-tool problems.