melchizedek
Melchizedek is a headless Node.js CLI for running hierarchical agent graphs — called Syndicates — built on the Google Agent Development Kit. It’s the reference implementation for many of the orchestration ideas discussed in the posts: delegation, critic loops, hierarchical task decomposition, and long-term memory.
what it does
- YAML-defined orchestration — an entire multi-agent hierarchy (models, prompts, tools, delegation rules) is declared in a single YAML file. Swap a subagent’s model with one line.
- Multi-model — Gemini natively, Claude via a custom LLM adapter. Each agent in a syndicate can run on a different provider.
- Persistent sessions & memory — Supabase-backed session persistence and pgvector long-term memory, with composite session keys so parallel subagents never corrupt each other’s context.
- A2A service mode — any syndicate can be served as a stateless JSON-RPC agent-to-agent endpoint over HTTP, with bring-your-own-key routing so inference costs land on the caller.
why it matters for this curriculum
Most agent frameworks hide the orchestration inside code. Melchizedek makes the shape of a multi-agent system a readable artifact — you can diff two syndicates the way you diff two configs. For teaching, that’s the point: the architecture of delegation is on the page, not buried in a call stack.
Posts that use melchizedek as their worked example will link here, and this page will grow into the framework’s public home — walkthroughs, example syndicates, and eventually live demos served from its A2A mode.
example syndicates
| Syndicate | Pattern it demonstrates |
|---|---|
| Global Synthesis Council | parallel research + synthesis |
| Delegation router | routing work to specialist subagents |
| Critic workflow | structured review loops |
| Hierarchical decomposition | breaking goals into subtasks |
| Financial Council | multi-analyst report generation |
Source code and full documentation will be linked here when the repository goes public.