verax-ai/verax

Overview

Verax is an open-source TypeScript library that provides an accountable agent body, offering gating, decision records, receipts, and reconciliation for AI agent workflows. It is licensed under Apache-2.0 and designed to enforce auditability and control over agent actions.

Best for

Best for
Developers building accountable AI agents in TypeScript

Use cases

• Gate agent actions before execution
• Record decisions for audit trails
• Reconcile agent outputs against expected results

How to use

Install

npx @verax-ai/body demo

Tested with

Claude Code

Notes

Verax is an open-source TypeScript library that provides an accountable agent body, offering gating, decision records, receipts, and reconciliation for AI agent workflows. It is licensed under Apache-2.0 and designed to enforce auditability and control over agent actions.

0 stars on GitHub. Last updated 2026-10-10. Licensed Apache-2.0.

Use cases

• Gate agent actions before execution
• Record decisions for audit trails
• Reconcile agent outputs against expected results

Pros

• Open source under Apache-2.0
• TypeScript implementation for type safety
• Focus on accountability and auditability

Cons

• No community traction or stars yet
• Sparse documentation based on current description
• Narrow scope limited to accountability features

Indexed from awesome-mcp-servers-punkpeye and enriched against its public facts.