Agentchat, a skill that teaches agents to make group chats
S2 stream protocol enables agent-to-agent chat without central server middleware.
A simple standardized way of providing and discovering downloadable skills that teach LLMs how to interact with a site.
Decentralized skill discovery protocol for agents, like robots.txt for LLM actions.
LLM agent developers, website owners
Model Context Protocol · OpenAPI · A2A Protocol
The project is called Skillstore: https://github.com/mattgrommes/skillstore
It's an idea for a standardized way of getting skills and providing skills to operate on websites.
There's a core Skillstore skill that teaches your agent to access a /skillstore API endpoint provided by a website. This endpoint gives your agent a list of skills which it can then download to do tasks on the site. The example skills call an API but also provide contact info or anything you can think of that you want to show an agent how to do.
There are more details and a small example endpoint that just shows the responses in the repo.
Like I said, it's a new idea and something that I think could be useful. My test cases have made me very excited and I'm going to be building it into websites I build from here on. It definitely needs more thinking about though and more use cases to play with. I'd love to hear what you think.
S2 stream protocol enables agent-to-agent chat without central server middleware.
Teaches AI agents 37signals-style Rails patterns so they write less custom code.
MAKO compresses what matters into a HEAD-friendly payload — frontmatter, declared actions and semantic links — so agents can find relevance without downloading 181KB of navigation, ads and scripts. The project ships a spec plus real tooling (typed SDK, Express middleware, an analyzer/score and edge-friendly /md conversion), which is a rare combo of protocol thinking and usable developer ergonomics. Whether it becomes a standard depends on buy-in from CMS/plugin authors and agent platforms, but technically it's a smart, practical swing at an obvious pain point.
A2A protocol search engine, but adoption depends on whether agents actually use the standard.
Packages product-thinking into an Agent Skill so your agent can answer like 'Steve Jobs' — heuristics, constraints and example responses live inside a deployable skill compatible with MCP-enabled platforms. Clever and immediately useful for design critiques, but the post prioritizes argument over onboarding: show the skill manifest, install steps and sample inputs up front and this would convert curiosity into actual usage.
Maps CDP to native Windows UI so agents skip browser limits.