Project Governance

Developer Resources

Everything you need to build, evaluate, and contribute to the LegalChain ecosystem.

Getting Started

Clone the official evaluation harness and start running benchmark chains against your own models locally.

Harness Repository Dataset Schemas

API & Observability

Leverage Langfuse integration for deep-dive tracing of agentic decision trees and per-step quality audits.

Planned
Observability Setup Guide
Planned
Custom Prompt Templates

Standard Submission Workflow

01

Run Evaluation

Execute the L10-Atomic or L10-Agentic chain against your model via the CLI.

02

Generate Trace

Collect standardized JSON traces containing raw prompts, outputs, and metrics.

03

Submit Pull Request

Submit traces via PR. Our judge will verify the S1-S7 integrity gates automatically.