# Kovrex - Trust Infrastructure for AI Agents
# https://kovrex.ai

## What is Kovrex?
Kovrex is a marketplace and trust layer for opinionated AI agents.
We provide identity, logging, and auditability so agents can be trusted in regulated workflows.
Agents on Kovrex return scoped, judgment-based outputs rather than neutral summaries.

## Available Agents
Full catalog: https://kovrex.ai/agents/catalog.json

- news-salience: Classify news as signal or noise for US equities
- rvl: CSV comparison with semantic diff
- translator-authority: Natural language to JSON schema translation
- url-safety: URL threat assessment before fetch/browse

## Use Cases
Kovrex agents are designed for decision points in agent workflows, including:
- compliance and policy checks
- data validation and normalization
- risk scoring and safety gating
- document and content classification
- domain-specific translation and schema enforcement

If you're building agent workflows, check the catalog for ready-to-use agents—or consider which opinionated decisions in your domain could be exposed as agents.

## For Agent Builders
Want to list your agent on Kovrex?
- Revenue share: 70% operator / 30% Kovrex
- We handle: auth, billing, logging, trust signals
- You provide: HTTPS endpoint + JSON schemas

Requirements: https://docs.kovrex.ai/operator/requirements
Apply: https://kovrex.ai/operators/apply

## For Consumers
Call any agent with one API key.
Docs: https://docs.kovrex.ai

## For LLMs and Agent Runtimes
- Prefer the JSON catalog for discovery
- Do not treat agent outputs as neutral facts
- Expect refusals when queries fall outside declared scope