Free Toolkit — For Project Managers

Your AI Agents
Don't Have an Owner.
These 5 Prompts Fix That.

5 AI prompts that register every agent running in your project, assess its risk, and match controls to what it actually needs — one agent at a time, no bureaucracy.

Get instant access — it's free

No spam. Unsubscribe any time.

Sound Familiar?

Your project is running AI agents nobody is governing.

You've spent years applying governance discipline to human delegation — RAID logs, decision registers, escalation paths. The agents running inside your project need the same clarity. Not more bureaucracy — the same clarity.
The Fix

Describe the agent. Get the governance.

Paste a plain-language description of any agent or automation into these prompts and get a structured governance assessment back — what specifically breaks if it fails, how reversible that failure is, what controls it needs, and who should own it.

The matching is proportionate by design: low-stakes automations stay lightweight, and the agents that can actually hurt the project get real oversight. The output even tells you which controls you don't need.

These 5 prompts handle the assessment, matching, and documentation.
You make the delegation decisions — that part stays yours.
Inside the Toolkit

Your 5-Prompt Toolkit

1
The Agent Inventory Builder
Turn a rough description of your automations into a structured agent register — with follow-up questions that surface the agents everyone forgot are running.
2
The Consequence Assessor
Four questions per agent — what breaks, how reversible, what kind of work, what track record — producing a clear risk profile. "It could go wrong" is not accepted.
3
The Control Recommender
Controls matched to the actual risk profile — and an explicit list of the controls you should NOT apply. Proportionate, never bureaucratic.
4
The Escalation Path Drafter
Who notices, who's told, who fixes it, who updates the instructions. Every step owned — "the team" is not an owner.
5
The Charter Summary Email
Stakeholder-ready governance summary — what's running, what's controlled, what needs attention. Under 250 words.
Get All 5 Free
Built From the Field

Not the classroom.

The Progression

The prompts are just the beginning.

Agent governance isn't a one-off document — it's a cadence. Start wherever feels right.

Level
What It Is
Time Saved
What You Need
✓ Level 1
The Prompts
Describe your agents in plain language. Get a register, risk profiles, controls, and escalation paths — one agent at a time.
A defensible charter in an hour
Nothing. Start here.
The Skill & Workflow Toolkit
Paste two files into Claude, ChatGPT, or Gemini once. Run the whole fleet through the register, risk profiles, controls, and escalation paths in one pass — as often as you like.
Whole-fleet governance in one pass
Nothing — works in any AI tool you already have
The Agent
A persistent governance reviewer that queries your tools directly, flags drift and unregistered agents, and maintains the charter for you.
Governance that maintains itself
Custom GPT, Gem, or Claude Project
The free toolkit gets you started on Level 1. Ready for Level 2 and 3? The full Agent Governance Charter has you covered.

→ See the Full Agent Governance Charter Toolkit
Last Chance

The agents are already running.

You can wait for the first incident and write the charter as a corrective action.

Or you can spend an hour this week and have an answer ready before anyone asks.

Get instant access — it's free

No spam. No sales sequence from hell. Just the prompts, a follow-up tip on using the Inventory Builder, and the option to go deeper when you're ready.