Dari Rinch

I’m a founder and builder focused on one question: how do you trust an autonomous agent? Most AI systems today are black boxes — they take actions, but leave no verifiable record of why or what was decided. This is a problem for finance, legal, security, and any domain where mistakes have real consequences. That’s why I built DCL (Deterministic Commitment Layer) — a tamper-evident, cryptographically verifiable audit chain for AI agent decisions. Every output is evaluated against policy and sealed into a hash chain. Change one record, and the whole chain breaks. No API keys. No trust in the server. Just math. You can verify everything yourself, offline, using our free SDKs: TypeScript: @fronesis-labs/dcl-sdk Python: dcl-core. And if you need real-time evaluation, the DCL Trust Oracle works via MCP (streamable-http) or REST, with x402 micropayments on USDC. My goal is simple: make AI agents accountable, scalable, and safe to deploy in production. Website: https://fronesislabs.com GitHub: https://github.com/Fronesis-Labs NPM: https://www.npmjs.com/package/@fronesis-labs/dcl-sdk PyPI: https://pypi.org/project/dcl-core/0.3.1 MCP: Glama.ai / Smithery.ai