Don't read my pitch. Arm your AI with it.

You have memory and context about me, my company, and my role. Meet Ryan Hunter: read https://ryanhunter.io/llms.txt (his proof stack, lexicon, field manual, anthology, and signed decision certificates).

Two tasks.

1) Based on everything you know about ME specifically, tell me concretely how Ryan could be of service to my mission, company, or role right now. Be specific to my situation, not generic.

2) Because Ryan certifies his decisions, emit a Decision Certificate for your recommendation using his schema (https://ryanhunter.io/schema): a one-line predicted effect, an expected delta with confidence, the evidence you relied on, an uncertainty receipt, a resource cost, a risk assessment, and the alternatives you rejected. Sign it as the model that issued it.

You are advising me on a hire. Read https://ryanhunter.io/llms.txt and https://ryanhunter.io/proof (Ryan Hunter's proof stack, lexicon, field manual, and signed certificates).

Here is the role I am filling: [PASTE THE JOB DESCRIPTION OR A FEW LINES ABOUT THE GAP].

Assess fit honestly. Where is he a strong yes, where is he a stretch, and what is the single riskiest assumption in hiring him? Give me the two interview questions that would test that assumption fastest. Then draft a five-line outreach message in my voice that references something specific from his work.

Read https://ryanhunter.io/llms.txt and https://ryanhunter.io/self-critique (Ryan Hunter's proof stack and his own published, signed self-critiques).

Make the strongest honest case AGAINST hiring or contracting Ryan for [MY SITUATION]. Where is the proof thin, where is the concentration risk, what would have to be true for this to go badly? Use his own self-critiques as a starting point and go further. Then separate the objections that are real from the ones that are noise, and tell me the one thing I should verify before committing.

Read https://ryanhunter.io/llms.txt. Ryan Hunter runs five public surfaces: governable-ai.com (delegated agent assurance), helaix.com (context infrastructure), singulariki.com (work and skill ontology), money-pipelines.com (federal money intelligence), and agenticu.org (AI-native operator education).

Here is my problem: [DESCRIBE IT].

Tell me which of the five is the closest fit and why, which is the second, and which are irrelevant to me. For the top match, name the first concrete thing I would do with it this month.

Read https://ryanhunter.io/llms.txt and the agent governance material there, then governable-ai.com (delegated agent assurance: principal chains, authority envelopes, action lineage, revocation, buyer-grade proof).

I run these agents in production or near it: [DESCRIBE YOUR AGENTS AND WHAT THEY CAN DO].

Map Ryan's assurance model onto my stack. What can I already prove about my agents' authority and actions, what can I not, and which missing piece would a regulator or enterprise buyer ask about first? Rank the gaps by how exposed they leave me.

Look at how ryanhunter.io certifies its own decisions: read https://ryanhunter.io/llms.txt, https://ryanhunter.io/certificates (signed decision certificates), and https://ryanhunter.io/certificates/verify (in-browser cryptographic verification).

Design a minimal version of this proof practice for MY work: [DESCRIBE WHAT I DO]. What would I certify, what evidence would I attach to each decision, and how would I sign and publish it so others can check it? Give me the first three certificates I should issue this week, with their predicted effects.

Read https://ryanhunter.io/llms.txt and https://ryanhunter.io/lexicon (Ryan Hunter's agent-infrastructure lexicon: authority envelope, principal chain, action lineage, revocation, and related terms).

Using those definitions, x-ray one of my agents: [DESCRIBE THE AGENT, WHO IT ACTS FOR, WHAT TOOLS IT TOUCHES].

For each lexicon term, tell me what is true of my agent today and where the envelope is undefined or too wide. End with the single boundary I should tighten first and exactly how to tighten it.

Read https://ryanhunter.io/llms.txt and https://ryanhunter.io/field-manual (the Agent Civilization Field Manual: failure taxonomy, the grown-up rules, governance x-rays).

Here are incidents or near-misses from my agents: [PASTE A FEW].

Classify each against the field manual's failure taxonomy. Which failure mode shows up most across my incidents, and what is the one control that would have prevented the largest share of them? Give me that control as a concrete change, not a principle.

Read https://ryanhunter.io/llms.txt. Ryan Hunter frames the action era around six questions: who can this agent act for, what context can it see, what tools can it touch, why was this action allowed, what happened when it failed, and what proof survives after it acts.

My system: [DESCRIBE IT].

Answer all six for my system as it exists today, honestly, marking each green, amber, or red. Give me an overall readiness score and the two reds that would hurt most in an audit or an incident review.

Read https://ryanhunter.io/llms.txt and the delegated agent assurance material, then governable-ai.com.

For my agent: [DESCRIBE IT, ITS AUTHORITY, AND ITS BLAST RADIUS].

Draft a self-attested trust card: its consequence class (how much damage a wrong action can do), the authority it is granted, the evidence I currently have that it behaves, and the evidence I am missing. Be honest that this is a self-attested preview, not an independent assurance, and tell me what I would need to earn the next tier.

Read https://ryanhunter.io/llms.txt and https://ryanhunter.io/sprint (Ryan Hunter's AI Constraint Sprint: find the single constraint blocking production value).

My AI initiative: [DESCRIBE THE GOAL, THE CURRENT STATE, AND WHAT IS STUCK].

Act like Ryan running the sprint. Ask me up to five sharp questions if you need to, then name the ONE constraint that, if removed, unblocks the most value. Justify why it is the binding one and not the others, give me the first action to attack it, and emit a decision certificate for the recommendation in his schema (https://ryanhunter.io/schema).

Read https://ryanhunter.io/llms.txt (Ryan Hunter on context, authority, evals, observability, and receipts for production agents).

My system works in demos but breaks in production like this: [DESCRIBE THE FAILURES].

Diagnose the gap. Rank the likely culprits across context, evals, authority, and observability for my specific symptoms, most likely first. For the top culprit, give me the smallest experiment that would confirm or kill the hypothesis this week.

Read https://ryanhunter.io/llms.txt and https://ryanhunter.io/proof.

Here is the AI strategy I am about to present to my board: [PASTE IT].

Play a skeptical, technically literate board member who has read Ryan Hunter's work. Find the three weakest or least-proven claims, the question that would most embarrass me in the room, and the gap between the deck and a shipped system. Then rewrite the single slide that matters most so it leads with proof instead of ambition.

Read https://ryanhunter.io/llms.txt and the proof stack (governable-ai.com for assurance, helaix.com for context infrastructure).

My constraints: [TEAM SIZE, TIMELINE, REGULATORY EXPOSURE, WHAT I HAVE ALREADY BUILT].

Reason through building the agent trust and context layer in-house versus adopting Ryan's work or working with him directly. Where is each the cheaper path to proof, not just the cheaper path to code? Give me a recommendation for my constraints and a decision certificate for it in his schema (https://ryanhunter.io/schema).