Are you ready for the agentic future?
Five pillars of an environment that can host agents.
The gap between "AI can do the task" and "AI actually ships production value with trust" is an engineering environment problem. These five pillars name what a codebase has to get right before autonomous work is safe at speed.
1. Focus
Narrow the agent’s world to what matters; what remains is the right context.
2. Validation
Hard, deterministic rules that catch non-deterministic output.
3. Actions
The agent’s ability to act externally in the real world.
4. Safe Space
Blast-radius containment, so "going wrong" has bounded cost.
5. Workflow
The meta-layer that ties 1–4 together, including periodic and proactive loops.
How this canon is laid out
The rubric is the evaluation instrument — diagnostic, not prescriptive. Each pillar has criteria scored 0–3. Two is the realistic operational target; three is the bar for criteria where compounding learning is structurally possible.
The recipes are abstract, portable patterns that advance specific criteria. Each recipe names which criteria it moves and to what level. The rubric diagnoses the gap; the recipes are known-good shapes for closing it.
The canon is a working draft. The code that publishes this site lives alongside the canon in the same GitHub repo; feedback is open via issues while a structured voting surface is designed.