Audit conformal interval coverage

Use this when your model emits prediction intervals or confidence bands and you want to audit whether realised outcomes actually fall inside them at the claimed

liveverifiedunknown~5s typical

mcpprobedmcp

$0.0020

per call

plus $0.001 platform fee per run

Open MCP server; Cracked bills routing only

Provider Upstream docs

$Set up https://cracked-ai-pearl.vercel.app/SKILL.md, then use Cracked to run mcp-com-alphaassay-mcp /assay_conformal for me.

Input

Field	Type	Description
intervals_lower*	array	Lower bounds of your prediction intervals, per point.
intervals_upper*	array	Upper bounds of your prediction intervals, per point.
outcomes*	array	Realised outcomes each interval was meant to cover.
claimed_coverage	string	The coverage level you claim, e.g. 0.9 for 90% prediction intervals. default 0.9
calibration_size	string	Optional size of the calibration set used to build the intervals; enables the exact beta-b
api_key	string	API key for a paid MCP check. Hosted Streamable HTTP clients should send it in the Authori default ""
request_id	string	Optional idempotency key for this paid execution. Retry the same request_id with the same

Call it

curl

curl https://cracked-ai-pearl.vercel.app/v1/run \
  -H "Authorization: Bearer ck_live_..." -H "content-type: application/json" \
  -d '{"provider":"mcp-com-alphaassay-mcp","endpoint":"/assay_conformal","input":{"intervals_lower":[],"intervals_upper":[],"outcomes":[],"claimed_coverage":0.9,"api_key":""}}'

cli

npx cracked-ai run -p mcp-com-alphaassay-mcp -e /assay_conformal -i '{"intervals_lower":[],"intervals_upper":[],"outcomes":[],"claimed_coverage":0.9,"api_key":""}'

mcp

run_tool({ provider: "mcp-com-alphaassay-mcp", endpoint: "/assay_conformal", input: {"intervals_lower":[],"intervals_upper":[],"outcomes":[],"claimed_coverage":0.9,"api_key":""} })

Try it

Runs against your signed-in workspace balance. Sign in if you have not.