2026-05-24

Denoise

The ecosystem for AI coding agents is standardizing, with a surge in plugins, skills, and orchestration tools built on a few core platforms.

Today's trends show a clear shift from building novel AI agents to creating the tooling layer that makes existing platforms like Claude Code useful.

2026-05-242026-05-24T13:51:52Zrules v1Healthyrepos 43signals 43

Top 3 changes

  • anthropics/claude-plugins-official shows the platformization of agent skills, with first-party players establishing official directories.
  • Lum1104/Understand-Anything shows a move from unstructured context fetching to structured knowledge graphs for more capable agents.
  • multica-ai/multica shows the evolution from single-agent scripts to managed platforms for coordinating teams of agents.

Strategic insights

#01The "Agent Skill" is emerging as a standard software primitive. Repos like mattpocock/skills, addyosmani/agent-skills, and dotnet/skills show a pattern of packaging agent capabilities into reusable, shareable formats, creating a new layer of the software stack.
#02The market is consolidating around a few agent runtimes. The frequent mention of Claude Code, Codex, and Cursor across dozens of repos indicates that the developer ecosystem is converging on these platforms, building tools for them rather than competing with them.
#03Orchestration is the new frontier. With single agents becoming a commodity, the complexity has moved to managing multi-agent systems. Projects like multica-ai/multica, ruvnet/ruflo, and mattpocock/sandcastle are building the infrastructure to coordinate swarms of agents for complex tasks.
#04Access-as-a-feature drives trends. A recurring pattern is tools like Alishahryar1/free-claude-code and decolua/9router that provide free or simplified access to paid AI models. Their popularity is often tied more to circumventing costs than to technical innovation.

Trend tracking

Today

A wave of new projects surged, led by the tutorial collection `codecrafters-io/build-your-own-x` alongside specialized tools like the chess engine `666ghj/MiroFish` and P2P OS `earendil-works/pi`.

6 Strengthening · 2 Continuous · 8 First seen
Expand list (16)
ProjectYesterday ⭐Today ⭐ΔState
codecrafters-io/build-your-own-x504k🆕First seen
666ghj/MiroFish62k🆕First seen
earendil-works/pi54k🆕First seen
blakeblackshear/frigate33k🆕First seen
Alishahryar1/free-claude-code29k🆕First seen
shiyu-coder/Kronos26k🆕First seen
manaflow-ai/cmux19k🆕First seen
anthropics/knowledge-work-plugins13k🆕First seen
Lum1104/Understand-Anything20k24k+4040Strengthening
colbymchenry/codegraph18k21k+2935Strengthening
multica-ai/andrej-karpathy-skills149k151k+2467Continuous
rohitg00/ai-engineering-from-scratch13k15k+1876Strengthening
anthropics/claude-plugins-official26k27k+1091Strengthening
mukul975/Anthropic-Cybersecurity-Skills7.0k7.9k+923Strengthening
multica-ai/multica32k32k+569Continuous
dotnet/skills2.7k2.9k+186Strengthening

This week

AI-related tools and educational repos show persistent growth, with `HKUDS/CLI-Anything` debuting strongly alongside established gainers like `colbymchenry/codegraph` and `Lum1104/Understand-Anything`.

8 Strengthening · 4 Continuous · 1 Stable · 4 First seen
Expand list (17)
ProjectYesterday ⭐Today ⭐ΔState
HKUDS/CLI-Anything40k🆕First seen
yikart/AiToEarn16k🆕First seen
Lum1104/Understand-Anything20k24k+4040Strengthening
stablyai/orca3.2k🆕First seen
colbymchenry/codegraph18k21k+2935Strengthening
rohitg00/ai-engineering-from-scratch13k15k+1875Strengthening
obra/superpowers204k205k+1004Stable
Imbad0202/academic-research-skills19k20k+842Strengthening
CloakHQ/CloakBrowser19k20k+773Strengthening
tinyhumansai/openhuman26k27k+756Continuous
cursor/plugins711🆕First seen
ruvnet/RuView64k65k+661Continuous
rohitg00/agentmemory17k17k+519Strengthening
supertone-inc/supertonic9.7k10k+347Strengthening
can1357/oh-my-pi6.6k6.9k+323Strengthening
datawhalechina/easy-vibe14k14k+256Continuous
K-Dense-AI/scientific-agent-skills25k26k+251Continuous

This month

Interest is consolidating around established AI skill-building repos like `multica-ai/andrej-karpathy-skills`, while `colbymchenry/codegraph` demonstrates entrenched popularity across all timeframes.

5 Strengthening · 10 Continuous · 3 Stable · 2 First seen
Expand list (20)
ProjectYesterday ⭐Today ⭐ΔState
yikart/AiToEarn16k🆕First seen
CJackHwang/ds2api4.6k🆕First seen
colbymchenry/codegraph18k21k+2935Strengthening
multica-ai/andrej-karpathy-skills149k151k+2468Continuous
mattpocock/skills102k103k+1213Continuous
Imbad0202/academic-research-skills19k20k+842Strengthening
CloakHQ/CloakBrowser19k20k+773Strengthening
Alishahryar1/free-claude-code28k29k+530Continuous
rohitg00/agentmemory17k17k+519Strengthening
HKUDS/ViMax6.8k7.2k+352Strengthening
TauricResearch/TradingAgents79k79k+337Stable
ruvnet/ruflo54k55k+332Continuous
anthropics/financial-services27k27k+302Continuous
addyosmani/agent-skills45k45k+245Continuous
decolua/9router14k14k+244Continuous
ComposioHQ/awesome-codex-skills11k11k+235Continuous
AIDC-AI/Pixelle-Video19k19k+185Continuous
soxoj/maigret30k30k+140Stable
mattpocock/sandcastle4.8k4.9k+122Continuous
bytedance/UI-TARS-desktop35k35k+81Stable

Categories· 8

Memory, RAG & Context(3)

This category is moving beyond simple vector search. The trending tools, Understand-Anything and codegraph, focus on creating structured knowledge graphs from codebases, providing agents with deeper, more reliable context than raw text retrieval.

  • This tool generates interactive knowledge graphs from codebases for AI agents to query. It's trending because it represents a move towards more structured and reliable context for agents, an improvement over simple file-based retrieval common in RAG systems.

    rising⭐ 24k
  • Provides a pre-indexed knowledge graph of code to reduce token usage and tool calls for AI agents. Similar to Understand-Anything, its rise points to the need for efficient, structured context as agent interactions become more complex and costly.

    rising⭐ 21k
  • A library for providing persistent memory to AI agents. It's a foundational component in the agent stack, and its repeated trending status indicates the ongoing importance of solving the state and memory problem for autonomous agents to perform complex, multi-step tasks.

    repeated⭐ 17k

Developer Tools & Productivity(4)

AI agent support is becoming a native feature in core developer tools. Instead of being a separate application, agent interactions are being integrated directly into terminals (cmux) and educational platforms (build-your-own-x), reflecting a deeper workflow integration.

  • A learning resource for AI engineering. This is not a tool but an educational project. Its popularity reflects the immense demand from developers to acquire the new skills needed to build and operate systems in the emerging AI agent paradigm.

    rising⭐ 15k
  • A terminal emulator for macOS based on Ghostty, specifically designed with features for AI coding agents like notifications and vertical tabs. This shows the developer toolchain adapting to new workflows introduced by AI agents, moving beyond simple editor plugins.

    new⭐ 19k
  • A beginner-focused coding course from a Chinese open-source community. Its presence in trends is likely driven by community promotion and serves as a regional educational resource rather than a globally impactful new technology.

    new⭐ 14k
  • A well-known, long-standing project that curates tutorials for rebuilding popular technologies from scratch. Its recent trending status may be due to new additions or renewed interest in foundational learning as a counterbalance to the abstraction-heavy world of AI development.

    new⭐ 504k

Prompt & Skill Libraries(13)

This is the epicenter of the agent tooling boom. We see the formalization of an ecosystem with official directories like anthropics/claude-plugins-official appearing alongside practitioner-driven collections like mattpocock/skills. It's the "app store" moment for AI agent capabilities.

  • A collection of practical, reusable skills for AI coding agents, shared by a well-known developer. Its popularity stems from its practitioner-focused approach, offering concrete, tested prompts and scripts rather than a theoretical framework. It's an example of community-led ecosystem building.

    cooling⭐ 103k
  • This is Anthropic's official directory of plugins for Claude Code. Its appearance is a major signal of platform maturation, showing a move towards a curated, controlled ecosystem similar to an app store, which standardizes how agents acquire new capabilities.

    new⭐ 27k
  • A specialized skill library for Claude Code focused on the academic research workflow. This demonstrates the verticalization of agent skills, where developers are creating domain-specific toolsets for fields like research, finance, and cybersecurity.

    repeated⭐ 20k
  • An official skill set from Anthropic for financial services tasks. Like their other official plugin repos, this reinforces the strategy of building a first-party, trusted ecosystem for high-stakes domains, encouraging enterprise adoption of their agent platform.

    cooling⭐ 27k
  • Another official plugin repository from Anthropic, this one targeting knowledge workers using their 'Claude Cowork' product. This further solidifies their platform play, creating specific tooling for different user segments and products within their ecosystem.

    new⭐ 13k
  • A single markdown file encapsulating Andrej Karpathy's observations on LLM coding into a skill for Claude Code. Its popularity is driven by the authority of the source and the simplicity of its format, showing high demand for expert-derived best practices.

    repeated⭐ 151k
  • A curated list of skills for the Codex platform. As an 'awesome list', it represents an early, community-driven stage of ecosystem organization, which is now being complemented and perhaps superseded by more structured, official directories from platform owners.

    cooling⭐ 11k
  • A collection of agent skills for scientific and technical domains. It's another example of the trend towards creating specialized, vertical-specific skill sets, enabling agents to perform more competently in complex fields like engineering and finance.

    new⭐ 26k
  • A repository of skills from the .NET team to help agents work with C# and the .NET ecosystem. This is significant as it shows a major language platform (Microsoft's .NET) officially engaging with the AI agent ecosystem, creating tooling to improve agent performance on their platform.

    new⭐ 2.9k
  • A large, structured collection of cybersecurity skills mapped to industry frameworks. Its value lies in its structured, comprehensive nature, aiming to provide a standardized, reliable skill set for agents operating in the high-stakes security domain.

    new⭐ 7.9k
  • A repository collecting methods to monetize AI usage. Repositories with this theme often trend due to financial incentives rather than technical novelty. They typically consist of prompt collections and scripts, representing a recurring, often low-signal pattern on trending lists.

    repeated⭐ 16k
  • The official plugin specification and repository for the Cursor IDE. This mirrors the pattern seen with Anthropic, where a platform owner (Cursor) is defining a formal structure for extending its AI-native editor, fostering a controlled ecosystem around its tool.

    new⭐ 711
  • A collection of engineering-focused skills for coding agents from a prominent Google engineer. Like the mattpocock/skills repo, its traction comes from the author's reputation and the focus on production-grade, practical applications, signaling a demand for quality control in agent skills.

    cooling⭐ 45k

Security & Reverse Engineering(2)

These tools represent persistent, specialized needs outside the main AI agent trend. CloakBrowser addresses the ongoing arms race in bot detection, while maigret is a classic OSINT tool. Their presence shows that fundamental security challenges remain relevant.

  • A stealth version of Chromium designed to evade bot detection systems. This is a specialized tool for web automation and scraping. Its continued popularity reflects the persistent cat-and-mouse game between automation tools and anti-bot services.

    repeated⭐ 20k
  • An OSINT tool that collects information on a person by searching their username across thousands of sites. This is a classic tool in the security and privacy space. Its presence on the list is not tied to the AI agent trend but to its ongoing utility.

    cooling⭐ 30k

AI Coding Tools & Agents(4)

This category consists of user-facing applications and wrappers built on top of foundational models and infra. Tools range from providing free access (free-claude-code) to offering specialized interfaces for the terminal (oh-my-pi) or desktop (orca), showing how the agent stack is being productized.

  • A project providing free terminal and VSCode access to claude-code. Its popularity is driven by offering a free gateway to a powerful, typically paid model. This pattern of wrapping paid APIs for free access is a common driver of short-term trends.

    repeated⭐ 29k
  • An AI coding agent that runs directly in the terminal. It's a self-contained implementation focused on a terminal-native workflow, integrating with tools like LSP. This represents the developer-as-user focus, building tools for their own core environment.

    new⭐ 6.9k
  • A desktop and mobile application described as an 'ADE' (Agent Development Environment) for managing a fleet of agents. This project aims to provide a dedicated IDE for working with agents, signaling a need for specialized tooling as agent development becomes more complex.

    new⭐ 3.2k
  • A middleware in Go that adapts protocols to be compatible with the DeepSeek model API. This is a technical utility for developers in the Chinese AI ecosystem, solving a specific interoperability problem. Its relevance is mainly for users of that particular model.

    cooling⭐ 4.6k

AI Infra & Protocols(12)

This category provides the picks and shovels for the agent gold rush. The focus is on orchestration (multica, ruflo) and abstraction layers (9router) to manage the complexity of building with and coordinating multiple agents, signaling a maturation from single-agent experiments.

  • An open-source project aiming to create a 'personal AI super intelligence'. The ambition is high, but the repository is new. It represents the ongoing interest in private, powerful, all-in-one AI agents, though it's too early to assess its substance.

    new⭐ 27k
  • An AI agent toolkit that bundles a CLI, unified LLM API, and UI components. It represents an effort to create an integrated development environment for agent-based applications, abstracting away some of the boilerplate involved in connecting various pieces of the agent stack.

    new⭐ 54k
  • An open-source platform for managing coding agents as 'teammates'. This project is notable because it addresses the operational side of using agents, focusing on task assignment and progress tracking, moving beyond single-shot scripts to continuous, managed agent workflows.

    new⭐ 32k
  • A framework for using multi-agent LLMs in financial trading. This is a domain-specific application of the broader multi-agent trend, targeting a high-value vertical where automation is a primary driver. Its cooling state suggests it may have peaked earlier.

    cooling⭐ 79k
  • A self-described 'universal swarm intelligence engine'. The description is abstract. This project, from a Chinese developer, appears to be an attempt at a general-purpose prediction engine using multi-agent principles. Its substance requires deeper code review.

    new⭐ 62k
  • Presents itself as both a framework and a methodology for agentic software development. It's an attempt to formalize the best practices for building with agents, moving from ad-hoc prompting to a more structured, engineering-driven approach.

    new⭐ 205k
  • A reverse proxy that routes requests to free tiers of various AI models. Its popularity is driven by providing free access to otherwise paid services. This is a utility for cost-conscious developers and hobbyists, focused on access rather than novel agent capabilities.

    cooling⭐ 14k
  • An 'agentic video generation' framework that breaks down video creation into roles like director and screenwriter. This applies the multi-agent system pattern to a creative, multimodal domain, which is a departure from the code- and text-heavy focus of most other trending agent repos.

    cooling⭐ 7.2k
  • A project aiming to make all software 'agent-native' via the command line. This is a highly ambitious infrastructure play, attempting to create a universal adapter layer for agents to interact with any existing CLI tool, potentially unlocking vast new capabilities.

    new⭐ 40k
  • An orchestration platform for deploying multi-agent systems, specifically highlighting Claude. It competes in the growing space of agent management platforms like Multica, focusing on building and coordinating complex, autonomous workflows for enterprise use cases.

    cooling⭐ 55k
  • An open-source multimodal AI agent stack from ByteDance. As a contribution from a major tech company, it signals serious investment in agent infrastructure, particularly in the multimodal space which combines vision and other inputs.

    cooling⭐ 35k
  • A TypeScript library for orchestrating sandboxed coding agents. From the same author as 'skills', this tool addresses the security and safety aspects of running autonomous code-executing agents, a critical component for production deployment.

    cooling⭐ 4.9k

On-device & Multimodal AI(4)

This category highlights a parallel track of AI development focused on edge computing and novel sensor data. On-device TTS (supertonic) and WiFi-based spatial sensing (RuView) are distinct from the cloud-centric agent ecosystem, focusing on privacy and real-world interaction.

  • A system that uses WiFi signals for spatial intelligence and presence detection without cameras. This project is notable for its novel use of ambient signals for sensing, representing an alternative, privacy-preserving approach to smart environment monitoring.

    new⭐ 65k
  • An on-device, multilingual Text-to-Speech (TTS) engine running via ONNX. Its focus on speed and local execution contrasts with the cloud-heavy agent development trend, catering to applications that require low latency and data privacy.

    new⭐ 10k
  • A tool described as a 'fully automated short video engine'. This repository targets the lucrative content creation market, promising to automate video production. The technical substance often varies in such projects, which attract attention with promises of automation.

    cooling⭐ 19k
  • A popular open-source Network Video Recorder (NVR) with local object detection. This is a mature project in the home automation and security space. Its trending status indicates sustained interest in on-premise, privacy-focused AI applications.

    new⭐ 33k

ML & GPU Infrastructure(1)

The single entry, Kronos, points towards the trend of creating specialized, domain-specific foundation models. While the broader trend is horizontal agent tooling, this signals continued investment in vertical AI for high-value sectors like finance.

  • A foundation model specifically for financial markets. This is an example of vertical AI, where models are trained on specialized data for a high-value domain. It stands in contrast to the general-purpose agent tooling that dominates today's trends.

    new⭐ 26k

Recent reports

  • 2026-06-03

    generated 2026-06-03T17:05:52Z · rules v1 · repos 45 · signals 45

  • 2026-06-02

    generated 2026-06-02T16:49:13Z · rules v1 · repos 43 · signals 43

  • 2026-06-01

    generated 2026-06-01T17:57:29Z · rules v1 · repos 44 · signals 44

  • 2026-05-31

    generated 2026-05-31T14:03:33Z · rules v1 · repos 44 · signals 44

  • 2026-05-30

    generated 2026-05-30T13:57:40Z · rules v1 · repos 48 · signals 48

  • 2026-05-29

    generated 2026-05-29T16:02:16Z · rules v1 · repos 46 · signals 46

  • 2026-05-28

    generated 2026-05-28T16:19:16Z · rules v1 · repos 48 · signals 48

  • 2026-05-27

    generated 2026-05-27T16:04:14Z · rules v1 · repos 45 · signals 45

  • 2026-05-26

    generated 2026-05-26T16:01:07Z · rules v1 · repos 45 · signals 45

  • 2026-05-25

    generated 2026-05-25T15:24:58Z · rules v1 · repos 45 · signals 45

  • 2026-05-24 · current

    generated 2026-05-24T13:51:52Z · rules v1 · repos 43 · signals 43

  • 2026-05-23

    generated 2026-05-23T13:55:04Z · rules v1 · repos 42 · signals 42

  • 2026-05-22

    generated 2026-05-22T14:56:21Z · rules v1 · repos 40 · signals 40

  • 2026-05-21

    generated 2026-05-21T15:46:27Z · rules v1 · repos 41 · signals 41

  • 2026-05-20

    generated 2026-05-20T15:48:11Z · rules v1 · repos 36 · signals 36

  • 2026-05-19

    generated 2026-05-19T15:42:42Z · rules v1 · repos 37 · signals 37

  • 2026-05-18

    generated 2026-05-18T15:51:10Z · rules v1 · repos 39 · signals 39

  • 2026-05-17

    generated 2026-05-17T13:49:34Z · rules v1 · repos 45 · signals 45

  • 2026-05-16

    generated 2026-05-16T13:49:33Z · rules v1 · repos 34 · signals 34

  • 2026-05-15

    generated 2026-05-15T14:26:57Z · rules v1 · repos 38 · signals 38

  • 2026-05-14

    generated 2026-05-14T14:32:26Z · rules v1 · repos 42 · signals 42

  • 2026-05-13

    generated 2026-05-13T14:53:18Z · rules v1 · repos 44 · signals 44

  • 2026-05-12

    generated 2026-05-12T14:42:52Z · rules v1 · repos 36 · signals 36

  • 2026-05-11 · fallback

    generated 2026-05-11T15:30:34Z · rules v1 · repos 40 · signals 40

  • 2026-05-10

    generated 2026-05-10T13:46:39Z · rules v1 · repos 37 · signals 37

  • 2026-05-09

    generated 2026-05-09T13:43:42Z · rules v1 · repos 40 · signals 40

  • 2026-05-08

    generated 2026-05-08T14:08:16Z · rules v1 · repos 38 · signals 38

  • 2026-05-07

    generated 2026-05-07T14:38:09Z · rules v1 · repos 37 · signals 37

  • 2026-05-06

    generated 2026-05-06T14:34:27Z · rules v1 · repos 36 · signals 36

  • 2026-05-04

    generated 2026-05-04T14:22:31Z · rules v1 · repos 36 · signals 36

  • 2026-05-03

    generated 2026-05-03T13:40:38Z · rules v1 · repos 29 · signals 29

  • 2026-05-02 · fallback

    generated 2026-05-02T13:39:14Z · rules v1 · repos 30 · signals 30

  • 2026-05-01 · fallback

    generated 2026-05-01T13:46:05Z · rules v1 · repos 32 · signals 32

  • 2026-04-30

    generated 2026-04-30T14:19:05Z · rules v1 · repos 36 · signals 36

  • 2026-04-29

    generated 2026-04-29T14:23:30Z · rules v1 · repos 36 · signals 36

  • 2026-04-28

    generated 2026-04-28T14:34:52Z · rules v1 · repos 37 · signals 37

  • 2026-04-27

    generated 2026-04-27T14:20:15Z · rules v1 · repos 38 · signals 38

  • 2026-04-26

    generated 2026-04-26T13:24:10Z · rules v1 · repos 37 · signals 37

  • 2026-04-25

    generated 2026-04-25T13:22:57Z · rules v1 · repos 41 · signals 41

  • 2026-04-24

    generated 2026-04-24T13:58:04Z · rules v1 · repos 40 · signals 40

  • 2026-04-23

    generated 2026-04-23T14:04:35Z · rules v1 · repos 43 · signals 43

  • 2026-04-22

    generated 2026-04-22T14:02:49Z · rules v1 · repos 41 · signals 41

  • 2026-04-21

    generated 2026-04-22T07:58:11Z · rules v1 · repos 40 · signals 40

  • 2026-04-20

    generated 2026-04-21T09:23:28Z · rules v1 · repos 40 · signals 40

  • 2026-04-19

    generated 2026-04-19T08:51:11Z · rules v1 · repos 36 · signals 36

  • 2026-04-18

    generated 2026-04-18T08:45:02Z · rules v1 · repos 39 · signals 39

  • 2026-04-17

    generated 2026-04-17T09:14:47Z · rules v1 · repos 40 · signals 40

  • 2026-04-03

    generated 2026-05-13T03:24:00Z · rules v1 · repos 26 · signals 26

  • 2026-04-02 · degraded

    generated 2026-04-03T19:29:39Z · rules v1 · repos 26 · signals 26

  • 2026-04-01 · fallback

    generated 2026-04-03T19:29:39Z · rules v1 · repos 26 · signals 26