Validate Agent Architecture (Consensus Mode)

Pro/Teams — N-shot CONSENSUS doctrine review of agentic code. ON CLIENT TIMEOUT — DO NOT RETRY THIS TOOL. Long-running (~80-120s for N=3 parallel LLM calls); MC

liveverifiedunknown~5s typical

mcpprobedaidesignblueprint

$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-aidesignblueprint-blueprint /architect.validate_consensus for me.

Input

Field	Type	Description
implementation_context*	string	The artifact under review. SEND FULL FILE CONTENTS VERBATIM — same constraint as architect
focus_area	string	Optional: narrow the review to a principle cluster or slug.
task	string	What the agent or workflow is trying to accomplish.
language	string	Programming language of the code (e.g. 'python').
repository	string	Iteration key. Consensus children all run unanchored (`private_session=True`), but the con
files	string	List of file paths relevant to the implementation.
goals	string	Specific safety or quality goals to evaluate against.
example_limit	integer	Max curated examples per child run. default 3
n	integer	Number of parallel child runs. Default 3 (the variance signal is visible at N=3; cost = 3× default 3

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-aidesignblueprint-blueprint","endpoint":"/architect.validate_consensus","input":{"implementation_context":"...","example_limit":3,"n":3}}'

cli

npx cracked-ai run -p mcp-com-aidesignblueprint-blueprint -e /architect.validate_consensus -i '{"implementation_context":"...","example_limit":3,"n":3}'

mcp

run_tool({ provider: "mcp-com-aidesignblueprint-blueprint", endpoint: "/architect.validate_consensus", input: {"implementation_context":"...","example_limit":3,"n":3} })

Try it

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