CAGE + ARCH

The two C4AIL practitioner protocols. CAGE sets the AI up before it does anything; ARCH verifies its work at every step. Together they form a Logic Pipe — structured, auditable AI work. These tools run entirely in your browser; nothing is sent anywhere. Paste the output into ChatGPT, Claude, or Gemini.

Centre for AI Leadership (C4AIL)

CAGE · set up

CAGE Builder

Fill in Context (all of it), Align (the plan-check), Goals, and Examples → copy a ready-to-paste block plus the "show me your plan first" prompt.

ARCH · verify

ARCH / Logic-Pipe Runner

Lay out your steps, run each through Action → Reasoning → Contextual Check → Horizon → export a documented Logic Pipe + a Decision-Survivability trace.

locate

Maturity × Autonomy

Six questions to find where you sit on the L0–6 × chat/agentic grid — and your single highest-leverage next move.

How to use it: Start with the CAGE Builder to brief the AI, then run multi-step work through the ARCH Runner. The golden rule: 30 seconds on Align saves 30 minutes unwinding the wrong artefact. You CAGE-initialise your agents and ARCH-verify your chats.