GraphTheory

Building installable Hermes Agent profiles, agent tooling, and specialist AI workflows.

I build open-source agent infrastructure for people who want practical, verifiable AI systems: profile distributions, reusable skills, catalog-ready agent identities, and domain specialists that can be installed, tested, and improved like normal software.

Featured original projects

• Hermes Profile Template: AI-friendly starter kit for building high quality, installable Hermes Agent profile distributions. Includes generators, validation, catalog snippets, CI, and release-ready structure.

• ContextForge RAG: installable Hermes Agent profile for production RAG architecture, AI agent workflows, evaluation, observability, and release discipline.

• ChainForge: security-first blockchain architect profile for smart contracts, protocol design, audits, tokenomics, and production-grade Web3 engineering.

What I care about

• Agent profiles that install cleanly with hermes profile install.
• Skills that encode reusable procedures, not one-off chat history.
• Validation, smoke tests, and visible provenance for every public agent package.
• Specialist agents with clear scope, refusal rules, output contracts, and verification discipline.
• RAG, blockchain, graph technology, and developer tooling that can survive real users.

Also building

• Solana Rug Guard: open-source trust and safety analysis for Solana tokens and wallets.

Start here

hermes profile install github.com/codegraphtheory/hermes-profile-template \
  --name profile-architect \
  --alias

profile-architect chat

Then ask it to create a specialist Hermes profile for your workflow.