Why we are building truval.dev
Software used to be built for people sitting at keyboards. A lot of new software is built for agents that read specs, call tools, and ship code without ever opening your marketing site. Those agents do not click “Get started.” They need a URL, a schema, and a key. If your product is a UI with an API on the side, you are invisible to that workflow.
That is the bet behind truval.dev: API infrastructure for AI agents. Not a platform that happens to expose an API, but endpoints that are designed to be consumed the way agents and IDE tooling consume software today—starting with a machine-readable contract, not a PDF.
One job per endpoint
We are intentionally boring in scope: each public route should do one thing and return a stable, documented shape. Our first surface is email verification (POST /v1/email/verify and friends: batch, stream, async). Every serious product collects email; agents wiring signups and automations need the same capability without owning SMTP plumbing, blocklists, and deliverability edge cases.
Agent-native, not agent-bolted-on
Being agent-native means more than “you can paste a curl into ChatGPT.” It means:
- OpenAPI 3.1 served at a stable URL so tools and codegen can treat it as source of truth.
- A remote MCP server so compatible clients can call verify and key management as tools without custom glue.
- llms.txt and llms-full.txt as deliberate entrypoints for models and assistants that start from plain text, not a browser.
The optional dashboard exists for humans (keys, billing, usage)—it does not gate API access. Agents and scripts talk to the same edge API as your production code.
Trust and defaults
Agents amplify whatever you plug in. We care about predictable pricing (no seats, clear caps and overages), fast paths on the verify API, and EU-hosted processing where it matters for privacy expectations—not because it is trendy, but because infrastructure choices are part of the contract with your users.
What comes next
Email verification is the first utility API. The same pattern—small surface, big clarity—will extend to other agent-sized jobs over time. If this resonates, grab a key at the dashboard, read the docs, and build something an agent can run end to end without a tour guide.