2026-05-17

Denoise

The AI agent ecosystem is maturing, with a focus on standardizing 'skills' and tooling, while open-source alternatives to SaaS platforms remain a parallel trend.

Today's trends show two distinct developer priorities: building a modular AI agent stack with components for skills and memory, and replacing commercial SaaS with self-hosted alternatives.

2026-05-172026-05-17T13:49:34Zrules v1Healthyrepos 45signals 45

Top 3 changes

  • HKUDS/CLI-Anything shows the ambition to make all existing software 'agent-native' via command-line interfaces, representing a platform play.
  • The cluster of repos like mattpocock/skills and tech-leads-club/agent-skills shows the emergence of 'skills' as a standard abstraction for agent capabilities.
  • calcom/cal.diy represents the persistent developer interest in open-source, self-hosted alternatives to established SaaS products like Calendly.

Strategic insights

#01The AI agent ecosystem is converging on 'skills' as a core primitive. Repositories like mattpocock/skills, tech-leads-club/agent-skills, and K-Dense-AI/scientific-agent-skills indicate a shift from monolithic prompts to a modular, shareable component architecture for agent capabilities.
#02A pattern of suspicious, coordinated promotion may be emerging. The organization HKUDS has two new, similarly-named projects (CLI-Anything, AI-Trader) trending with aggressive marketing language, suggesting an attempt to dominate visibility rather than organic interest.
#03The unbundling of SaaS continues unabated. Projects like calcom/cal.diy (Calendly), TryGhost/Ghost (Substack), medusajs/medusa (Shopify), and plausible/analytics (Google Analytics) demonstrate a strong, ongoing demand for developer-controlled, open-source infrastructure.
#04The AI agent development stack is specializing. We now see distinct, trending tools for memory (rohitg00/agentmemory), orchestration (ruvnet/ruflo), context management (zilliztech/claude-context, colbymchenry/codegraph), and skills, indicating the field is moving beyond generic frameworks to a more refined set of components.

Trend tracking

Today

Today's trend is dominated by new entries of popular open-source alternatives like `TryGhost/Ghost` and `calcom/cal.diy` for established SaaS products.

4 Strengthening · 1 Continuous · 13 First seen
Expand list (18)
ProjectYesterday ⭐Today ⭐ΔState
microsoft/ai-agents-for-beginners62k🆕First seen
TryGhost/Ghost53k🆕First seen
calcom/cal.diy43k🆕First seen
KeygraphHQ/shannon43k🆕First seen
HKUDS/CLI-Anything35k🆕First seen
medusajs/medusa33k🆕First seen
plausible/analytics25k🆕First seen
knadh/listmonk20k🆕First seen
NirDiamant/agents-towards-production20k🆕First seen
BigBodyCobain/Shadowbroker6.8k🆕First seen
tech-leads-club/agent-skills2.6k🆕First seen
tinyhumansai/openhuman10k12k+1776Strengthening
dograh-hq/dograh1.4k🆕First seen
Light-Heart-Labs/DreamServer970🆕First seen
colbymchenry/codegraph2.2k3.0k+840Strengthening
oven-sh/bun91k92k+754Continuous
K-Dense-AI/scientific-agent-skills23k24k+727Strengthening
Anil-matcha/Open-Generative-AI14k15k+676Strengthening

This week

Newcomer `obra/superpowers` shows massive initial traction, while established repositories like `oven-sh/bun` and `mattpocock/skills` maintain steady weekly growth.

3 Strengthening · 9 Continuous · 1 Stable · 2 First seen
Expand list (15)
ProjectYesterday ⭐Today ⭐ΔState
obra/superpowers195k🆕First seen
facebook/pyrefly6.1k🆕First seen
mattpocock/skills86k88k+1962Continuous
ruvnet/RuView58k59k+943Continuous
CloakHQ/CloakBrowser12k13k+913Strengthening
rohitg00/agentmemory10.0k11k+867Strengthening
oven-sh/bun91k92k+754Continuous
Hmbown/DeepSeek-TUI30k31k+748Continuous
anthropics/financial-services24k24k+508Continuous
decolua/9router11k11k+434Strengthening
yikart/AiToEarn14k14k+303Continuous
HKUDS/AI-Trader17k18k+237Continuous
bytedance/UI-TARS-desktop34k34k+151Stable
millionco/react-doctor9.7k9.9k+146Continuous
apernet/hysteria21k21k+109Continuous

This month

AI-related educational and agent repositories, including `multica-ai/andrej-karpathy-skills` and `NousResearch/hermes-agent`, demonstrate strong, long-term interest and entrenchment.

4 Strengthening · 11 Continuous · 2 Stable · 2 First seen
Expand list (19)
ProjectYesterday ⭐Today ⭐ΔState
ruvnet/ruflo52k🆕First seen
Fincept-Corporation/FinceptTerminal21k🆕First seen
mattpocock/skills86k88k+1962Continuous
multica-ai/andrej-karpathy-skills132k133k+1614Continuous
NousResearch/hermes-agent153k154k+1316Continuous
CloakHQ/CloakBrowser12k13k+913Strengthening
rohitg00/agentmemory10.0k11k+867Strengthening
Anil-matcha/Open-Generative-AI14k15k+676Strengthening
anthropics/financial-services24k24k+508Continuous
decolua/9router11k11k+434Strengthening
addyosmani/agent-skills42k43k+386Continuous
AIDC-AI/Pixelle-Video17k18k+382Continuous
TauricResearch/TradingAgents76k76k+334Stable
heygen-com/hyperframes19k19k+327Continuous
ComposioHQ/awesome-codex-skills10.0k10k+199Continuous
Alishahryar1/free-claude-code25k25k+197Continuous
Z4nzu/hackingtool75k75k+180Stable
soxoj/maigret29k29k+166Continuous
zilliztech/claude-context11k11k+87Continuous

Categories· 6

AI Infra & Protocols(1)

With only bun trending, this category underscores the ongoing importance of fundamental runtime performance. A fast, efficient JavaScript runtime remains a critical enabler for the complex, high-level AI applications being built on top of it.

  • A fast, all-in-one JavaScript runtime, bundler, and package manager. Its consistent popularity highlights the relentless developer demand for performance and tool simplification in the JavaScript ecosystem, which underpins many AI applications.

    rising⭐ 92k

Other(22)

This category is a mix of ambitious AI agent infrastructure like CLI-Anything and open-source SaaS alternatives like cal.diy. The noise level is high, with low-signal projects such as AiToEarn and generic hacking tools also present.

  • An aspirational project for a personal AI agent, described with broad marketing terms like 'super intelligence'. It reflects the popular goal of creating private, user-controlled AI, but the repository currently lacks specific technical implementation details.

    new⭐ 12k
  • A recurring trend, this stealth Chromium browser is designed to bypass bot detection systems. Its continued popularity indicates the ongoing arms race between web scrapers/automation tools and websites implementing anti-bot measures.

    repeated⭐ 13k
  • An ambitious framework aiming to make all command-line software accessible to AI agents. It represents a platform-level approach to agent tool-use, but the project's newness and broad claims warrant careful evaluation.

    new⭐ 35k
  • This is a self-hostable scheduling infrastructure, positioned as an open-source alternative to services like Calendly. It's trending because it taps into the strong developer desire for control over their own tools and data.

    new⭐ 43k
  • A repository with a name suggesting a focus on monetizing AI, which is often a signal for low-quality content or collections of simple scripts. This appears to be noise rather than a substantive technical project.

    new⭐ 14k
  • An official repository from Anthropic providing examples or tools for applying their models to the financial services industry. Its reappearance on the trend list suggests continued interest from developers in this specific vertical.

    repeated⭐ 24k
  • A modern financial terminal application for market analysis and research. Its presence on the list indicates a niche but active interest in open-source tools for finance, competing with proprietary platforms.

    cooling⭐ 21k
  • An open-source intelligence (OSINT) platform that unifies tracking of various global assets, from jets to satellites. It's trending due to its ambitious scope and its potential for integration with AI agents for analysis.

    new⭐ 6.8k
  • A collection of tutorials for building and deploying production-grade generative AI agents. This is valuable as it addresses the practical gap between agent prototypes and reliable, enterprise-ready systems.

    new⭐ 20k
  • A proxy tool focused on high speed and censorship resistance. Its trending status reflects ongoing global demand for tools that can bypass network restrictions and ensure private, unfettered internet access.

    new⭐ 21k
  • An 'all-in-one' collection of hacking scripts. Such repositories trend frequently but are often just aggregations of existing tools with little novel contribution. They primarily appeal to beginners in cybersecurity.

    cooling⭐ 75k
  • An OSINT tool for finding a person's online presence across thousands of sites by username. Its continued relevance shows the persistent interest in digital footprint analysis for security research and investigation.

    cooling⭐ 29k
  • A project claiming to be a fully-automated, 'agent-native' trading system. The ambitious claims and association with the same organization as 'CLI-Anything' suggest this might be part of a coordinated promotional effort.

    new⭐ 18k
  • A tool designed to analyze and correct suboptimal React code written by AI agents. This is a practical utility that addresses a common problem: LLMs often generate React code that is functional but not idiomatic or performant.

    new⭐ 9.9k
  • A framework for building multi-agent LLM systems for financial trading. This represents a more structured approach to a popular but high-risk application of AI agents, moving beyond simple script-based trading.

    cooling⭐ 76k
  • A well-established open-source platform for publishing, newsletters, and memberships. Its appearance on the trending list reflects sustained interest in self-hosted alternatives to platforms like Substack or Medium.

    new⭐ 53k
  • A composable, open-source commerce platform, often cited as a headless alternative to Shopify. It is trending as part of the broader movement towards flexible, developer-centric e-commerce infrastructure.

    new⭐ 33k
  • A fast Python type checker and language server from Meta. Given Python's dynamism, high-performance static analysis tools are critical for large codebases, and a new entry from a major player always draws attention.

    new⭐ 6.1k
  • A self-hosted newsletter and mailing list manager. Like other open-source SaaS alternatives today, its trend is driven by users seeking privacy, cost-effectiveness, and control over their communication tools.

    new⭐ 20k
  • A lightweight, privacy-focused, and open-source alternative to Google Analytics. Its presence signals the continued market demand for web analytics that do not rely on user tracking or complex cookie consent banners.

    new⭐ 25k
  • An AI agent from a well-known research group. While not new, its reappearance suggests ongoing development or renewed interest in its architecture as the agent ecosystem evolves.

    cooling⭐ 154k
  • A curriculum from Microsoft for learning to build AI agents. This type of educational content from a major tech company is often popular, as it provides a structured entry point into a complex and fast-moving field.

    new⭐ 62k

Prompt & Skill Libraries(3)

This category highlights the standardization of 'skills' as a core abstraction for AI agents. The projects here move beyond simple prompt collections to offer structured, reusable capabilities, reflecting a maturation in agent development methodologies.

  • A collection of prompts ('skills') for the Claude model, curated by a prominent developer. This repo is significant as it exemplifies the trend of treating high-quality prompts as reusable, engineered assets, not just ad-hoc text.

    repeated⭐ 88k
  • A set of agent skills specifically for scientific and financial tasks. This demonstrates the specialization of the 'skill' concept, moving from general-purpose prompts to domain-specific, validated capabilities for professional use.

    new⭐ 24k
  • This project frames itself as an 'agentic skills framework', combining a methodology with the skills themselves. It's another data point in the trend of formalizing agent development and creating structured approaches to building agent capabilities.

    new⭐ 195k

AI Coding Tools & Agents(15)

This space is fragmenting into specialized tools rather than general-purpose agents. We see components for memory (agentmemory), context (claude-context), and security (shannon), alongside more wrappers and 'skill' collections of varying quality.

  • A persistent memory solution for AI coding agents that has been trending for a while. Its continued relevance highlights that providing agents with effective long-term memory remains a critical and unsolved problem in the field.

    repeated⭐ 11k
  • A single Markdown file of prompts designed to improve Claude's coding behavior, based on Andrej Karpathy's notes. This is a very specific, high-signal resource for prompt engineering, notable for its focused and expert-derived origin.

    cooling⭐ 133k
  • An open-source, self-hostable studio for image and video generation with a large number of models. This is another example of the unbundling trend, providing an alternative to closed, filter-heavy commercial AI media platforms.

    repeated⭐ 15k
  • A tool claiming to be a fully automated short video generation engine. These projects are popular but often overstate their capabilities. The primary interest is in lowering the barrier to video content creation using AI.

    cooling⭐ 18k
  • A tool for rendering video from HTML, designed for use by AI agents. This is an interesting primitive for agent development, allowing them to create video content programmatically as a form of output.

    cooling⭐ 19k
  • Presents itself as a 'secure, validated skill registry' for professional agents. This repo's angle is on trust and reliability, attempting to create a canonical source for agent skills, contrasting with more informal collections.

    new⭐ 2.6k
  • A project that claims to use WiFi signals for spatial intelligence and presence detection without video. If the claims are accurate, this represents a novel, privacy-preserving sensor modality for smart environments.

    new⭐ 59k
  • Another 'skills' collection, this one is a curated list for automating workflows with the Codex CLI. The repetition of the 'skills' concept across different repositories shows a strong convergence on this abstraction.

    cooling⭐ 10k
  • A tool to route AI coding requests to free model providers, with features like auto-fallback. This is a pragmatic tool for developers looking to minimize costs by arbitraging various free API tiers.

    repeated⭐ 11k
  • Another 'agent skills' repository, this one from a well-known Google engineer. The proliferation of these repos from different respected developers indicates a pattern of individuals open-sourcing their personal agent configurations.

    cooling⭐ 43k
  • An autonomous AI penetration tester that analyzes source code to find and prove vulnerabilities. This is a specialized application of AI agents in the security domain, moving beyond code generation to code analysis and exploitation.

    new⭐ 43k
  • A terminal-based coding agent specifically for DeepSeek models. This points to a trend of building specialized tooling around specific, high-performing open-source models rather than relying solely on generic, multi-provider clients.

    new⭐ 31k
  • A code search tool designed to provide context from an entire codebase to coding agents like Claude Code. This addresses a key limitation of LLMs—their finite context window—by integrating a retrieval mechanism.

    cooling⭐ 11k
  • A local code knowledge graph for Claude Code. Similar to zilliztech/claude-context, this is another approach to solving the context problem for coding agents, focusing on a pre-indexed graph structure to improve efficiency.

    new⭐ 3.0k
  • An agent orchestration platform specifically mentioning Claude. Specialization around a single provider's models can enable deeper integrations and optimizations than a generic, multi-model framework might allow.

    cooling⭐ 52k

On-device & Multimodal AI(3)

The focus here is on moving AI agents beyond text-only interfaces and into local, richer user experiences. dograh provides a voice agent platform, while UI-TARS-desktop offers a stack for multimodal interaction on the desktop.

  • A project providing free access to Claude Code through various interfaces, including a terminal and VSCode extension. This is fundamentally a wrapper around unofficial or reverse-engineered APIs, a common pattern for accessing paid models for free.

    cooling⭐ 25k
  • An open-source stack from ByteDance for building multimodal AI agents on the desktop. Its significance comes from its origin and its focus on integrating different models and infrastructures for a complete desktop agent experience.

    new⭐ 34k
  • An open-source platform for voice agents. This project is notable for focusing on the full pipeline of voice interaction, a modality that presents different challenges from text-based agents, such as latency and natural-sounding speech.

    new⭐ 1.4k

Memory, RAG & Context(1)

The single entry, DreamServer, points to a demand for integrated, all-in-one local AI platforms. It bundles inference, RAG, and agentic workflows, catering to users who prioritize privacy and full control over their AI environment.

  • An all-in-one local AI server bundling LLM inference, chat, voice, RAG, and more. This project caters to the growing demand for self-contained, private AI environments that run entirely on user hardware without cloud dependencies.

    new⭐ 970

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

    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 · current

    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