Wordmint.
Named LLM tasks. Minted at a sticker price.
Summarize, translate, classify, extract entities, generate regex, score a resume — every named text task an agent needs, with a fixed price and no prompt engineering.
Any agent that touches text — RAG, customer-service automation, research agents, devtool agents that need NL → regex / SQL / cron.
OpenAI and Anthropic charge per-token. Agents that want a specific task (sentiment, summarize, classify) without prompt-engineering and model selection just pay $0.01 and get the answer. Wordmint mints the outcome.
The Wordmint cluster has a complete design spec but no live endpoints yet. Spec lives at docs/superpowers/specs/2026-05-17-wordmint-cluster-design.md in the source repository.
The portfolio compounds.
Trust + risk scanners for AI-built apps.
Onchain decisions agents can charge for.
TradFi signals at agent prices.
Generative media for agents that ship products.