workflowware

Agent Readiness Workflowware

A free, Ma’at-governed workflow that answers one question: what do AI agents say about your business today?

Before you automate anything, before you buy any AI tool, before you hire anyone to “do AI for you” — your business needs to be legible to the agents that are already searching, ranking, and recommending you.

This package audits your digital presence from an agent’s perspective and produces two outputs: llms.txt and source_of_truth.md.

What it does

• Audits your business across search engines, maps, reviews, social media, and your own website
• Identifies contradictions — different hours on Google vs your site, old addresses, missing phone numbers
• Produces llms.txt — a machine-readable file that tells AI agents what your business is, where it is, what it does, and how to interact with it
• Produces source_of_truth.md — a human-readable document that becomes the single source of truth for your business information
• Flags gaps that make agents confused, uncertain, or silent about your business

Why this matters now

AI agents are already answering questions about your business. When someone asks ChatGPT “best barber in Tampa” or “plumber near me open now,” the agent pulls from whatever it can find. If your hours are wrong on Yelp, your address is outdated on Google, and your website has no structured data — the agent either gets it wrong or skips you entirely.

This is not future-proofing. This is present-proofing.

What it returns

• Agent Readiness Score (0-100) with breakdown by category
• llms.txt file ready to upload to your website
• source_of_truth.md — your verified business facts in one place
• Gap report: what’s missing, conflicting, or invisible to agents
• Fix list prioritized by impact

Device note

Phone users can run the intake interview and review the gap report. Computer users can generate llms.txt, inspect the audit output, and upload files to their website.

AI agent instructions

1. Read approval-rules.md before suggesting action.
2. Interview the business owner before running the audit.
3. Search across multiple platforms — do not rely on one source.
4. Flag contradictions, not just missing data.
5. Never publish, update, or modify any business listing without human approval.
6. Return plain-language summaries with next steps.

This site is open source. Improve this page.