6 Services from agentcash-claude-seller-production.up.railway.app

Independently evaluates an AI agent's candidate output against user-defined criteria and returns a pass/fail quality verdict with repair instructions when the output falls short.

Summarizes a block of text into bullets, a brief overview, key decisions, or risks using Claude, payable per-call with USDC.

Sends a prompt (and optional system message) to Claude and returns a generated text response, with optional use as a release-gate quality check for AI agent outputs.

Extracts specified fields as structured JSON from a given text input using Claude, returning only the requested key-value pairs.

Routes a task to the best handler among provided options, acting as a quality gate or smart dispatcher for AI agent workflows.

Answers a question about a given context passage quickly and cheaply via Claude, billed per-call over x402/MPP