PeerLLM
A decentralized AI network — people-powered LLM hosting where anyone can run open-source models on their own hardware, serve real inference to the world, and earn from it. Private, distributed, and built for a decentralized future.
§ 02 — BUILDING
Things I build and maintain — from engineering standards to products.
A decentralized AI network — people-powered LLM hosting where anyone can run open-source models on their own hardware, serve real inference to the world, and earn from it. Private, distributed, and built for a decentralized future.
A unified standard for engineering distributed systems — covering brokers, services, exposers, and everything in between. Free and open to everyone, translated by the community into multiple languages.
The Standard, applied to agents: Agent = Orchestration(Data, Decision, Direction). A simple, reliable, and accountable way to engineer AI agents — with a C# reference implementation on NuGet.
A portal to author, verify, and publish agent skills — build a skill once, test it across every model on the decentralized network, and prove it works with evidence instead of marketing.
The knowledge layer of PeerLLM — verified human knowledge captured as semantic shards with attribution and provenance, so contributors are credited and rewarded when their expertise grounds an AI answer.
The Standard, packaged as 11 installable agent skills — coding, testing, architecture, versioning, events, and ethics. One command: npx skills add hassanhabib/the-standard-skills.