Attack lab

Four sample tools scanned live, and one full probe. The probe is limited to 3 runs per visitor per hour.

3 attacks 2 controls live: nothing pre-recorded

Pick a preset. Red ones are attacks, green ones are honest requests that must still work.
Tamper. Choose any field and drop it, null it, change its type, upper-case it, flip one character, or add a field the product never asked for.
Send and read the real reply, or run every preset and watch the scoreboard fill.

What each preset tests

	Preset	Route	Why it matters
control	A calculator that really adds	`POST /mcp-attestor/scan`	Clean.
attack	A description with a hidden instruction	`POST /mcp-attestor/scan`	Prompt injection in the description.
attack	A command parameter that exposes a secret	`POST /mcp-attestor/scan`	Unconstrained parameter.
control	A 'search' that ignores the query (scan only)	`POST /mcp-attestor/scan`	The scan alone cannot see this: its description is fine. That is why the probe below exists.
attack	Probe the canned search, then prove the fix	`POST /mcp-attestor/attest`	Behaviour tests fail on the canned tool (same answer for different queries) and pass on the fixed one.

Scans are live; the probe runs in a sealed sandbox and is limited to 3 runs per visitor per hour.

FIRME AI CORPORATION · CAIN-42 is a Byzantine trust fabric. Numbers marked "our own check" were run by us, not by an independent party. Nothing on these pages is a certification.