the whole ruleset, on one page
Five questions, asked by a machine.
No human reviews an application, so there is no one to persuade, no queue to jump, and no reason to charge you. What is on the left is the question in plain words. What is on the right is exactly how it is measured, so you can dispute it.
If those two words are new: MCP is the plug that lets an AI assistant use somebody else's software. A2A is the little card that software hands over to say what it is and who runs it. Neither has a referee. That is the gap this fills.
| 01 | Is it actually there, and does it do anything? | Responds to initialize and tools/list with at least one tool. |
| 02 | Does it say who it is, in a place a machine can read? | /.well-known/agent-card.json returns JSON with a name and a description. |
| 03 | Does it say who is paying it? | The card declares paid_by, referral_fee, listing_fee. Only silence disqualifies. We do not judge whether the answer is flattering. |
| 04 | Does it answer the same way twice? | Identical input returns identical output. Not measured unless you ask, because it means executing one of your tools, and the first tool a server lists may well be destructive. |
| 05 | Can you check our answer without trusting us? | Every verdict carries a record_sha256. Drop that field, stringify the rest in key order, hash it. It must match. |
Limits
What a green row does not mean
A badge that quietly means more than it measured is worse than no badge. So here is the ceiling, stated at the same size as the claim.
- ✗Not that any number it gives you is correct. This checks conduct and disclosure, not arithmetic. Checking the figures is a separate, paid layer.
- ✗Not that the disclosure is true. Missing disclosure disqualifies. A disclosure later shown to be false is grounds for removal, not something the machine can detect.
- ✗Not that the business behind it is any good. Competence is not measurable from an endpoint and we will not pretend otherwise.
- ✗Not a certificate. A row reflects a measurement taken at a time. It stops being green when the measurement does.
Three words the machine uses
Held, pending, verified
verified
Measured, and every condition that was measured was met.
pending
Measured, and a condition was not met. A statement about the server.
held
Not measured. A statement about our instrument. An instrument failure is not a statement about the thing it failed to measure, so it is never reported as one.