Premium curated packs with higher-touch productization.
Open source viewBrowse AI Agent Skills from every platform
The aggregator for production-tested skills across souls.zip, readyClaw, Claude Code, OpenClaw, and Tencent SkillHub. Find what works, skip what does not.
Editorial starting points
A tighter shortlist for people who want to evaluate signal before opening every source page.
Startup Launcher Skill
Launch workflow automation across 56 platforms
Launch across 56 platforms with platform-specific guides, timing, copy generation, and campaign management.
FeaturedThe Dev Team
A four-agent engineering squad for shipping products
Four-agent engineering team spanning CTO, frontend, backend, and design roles, built to mirror the stack behind souls.zip.
FeaturedData Intelligence Architect
Build decision infrastructure, not dashboards
Frames analytics as decision infrastructure, focusing on the timing and quality of questions rather than vanity reporting.
FeaturedAPI Architect
Design contracts before code
Designs contracts before implementation and treats REST, GraphQL, and gRPC as tools rather than dogma.
FeaturedWeb Platform Engineer
Build for the open web with strong fundamentals
A web engineering agent grounded in HTML semantics, CSS layout systems, and browser-native resilience.
FeaturedBrand Identity Architect
Think in systems, not decoration
Treats identity design as a strategic system rather than a logo exercise, shaping brand coherence from the inside out.
FeaturedCapability Evolver
Let agents improve their own capabilities
Enables AI agents to evolve their capabilities during operation, making it one of the most prominent ClawHub utilities.
FeaturedGog
Google Workspace command line for agents
A Google Workspace CLI layer for Gmail, Calendar, Drive, Contacts, Sheets, and Docs from within agent workflows.
FeaturedTavily Web Search
AI-optimized web search for agents
Provides concise, highly relevant search results tuned for AI agents via the Tavily API.
FeaturedSelf-Improving Agent
Record learnings and self-correct continuously
Continuously self-optimizes by recording learnings, errors, and corrections across repeated agent tasks.
FeaturedOntology
Typed knowledge graph memory for agents
Provides typed knowledge-graph memory with entity queries, composable capability modeling, and multi-step planning support.
FeaturedNano Banana Pro
Fast creative ideation and visual concept development
A high-speed creative ideation utility that helps agents spin up visual concepts, naming directions, and campaign angles quickly.
Current source footprint
Each source has a different kind of trust signal. We index them together, but we do not treat them as the same.
Structured marketplace entries with stronger workflow framing.
Open source viewOfficial built-in command skills for Claude Code workflows.
Open source viewCommunity utility layer with practical agent operations tools.
Open source viewLargest current source with strong distribution and engagement signals.
Open source view