6 articles
The reference companion to Everything Is a Prompt. Every mode of frao-cosaw-mcp, the message bus, the context monitor, the phase-orchestrator launch modes, the Docker wiring, and the exact self-injection sequence.
A model's output quality is set by the context it reasons over, not the wording of the prompt. How Frao engineers that context with COSAW — analysis, architecture, and implementation plan in stages, each phase launched as a dedicated, isolated, containerized agent.
The technical companion: five components, two namespaces (public vs. internal split-horizon), one writer, two nameservers in two locations. Architecture, configuration samples, measured change latency, security, and how we verified the cutover.
LLMs are pure intelligence with no hands. Agents are the hands, programmed to use that intelligence through a small set of primitive operations.
How a development workflow evolved through four generations of tooling, hit a hard wall at 350K tokens, and solved it with context-isolated sub-agents.
Environment Separation Plan: Development vs Production Status: Implemented and operational across all Frao stacks Approach: Dual VPN isolation ( frao.dev / frao.prod ) with same ports across both