Multi-agent scaffold for autonomous research workflows driven by swarm orchestration patterns.
R&D / Experimental. Maintained by Sage / 0SxD as part of an ongoing prompt-engineering and agent-skills research portfolio.
An agent architecture scaffold that explores how a swarm of specialized sub-agents can divide, coordinate, and accumulate research work across sessions. The design separates corpus intake, claim extraction, and skill drafting into discrete agent roles. This repo holds the governance files, corpus structure, and skill specs for that pattern - not a production system.
spec.md- session state and hot-cache protocol; read first per agent entrysystem_directive.md- agent operating directive and role constraintsbootstrap.md- boot sequence and session initialization instructionsskills.md- skill registry (all skills seeded; drafts in progress)metrics_ledger.md- cost and progress ledgercorpus/- organized reading corpus (3 sub-folders: agentic swarm orchestration, systems architecture, data structures)
MIT. See LICENSE.
Author: Sage / 0SxD
This repo is part of an active R&D portfolio. Content may move, change, or be withdrawn. Issues welcome but reviews are best-effort.