Fact-check claims and catch AI hallucinations with Lenz - verified, sourced verdicts inside your n8n workflows.

Provenance-aware evaluation and review-report primitives for Grayhaven Nerve.

Optional multi-model Arena capability for CodeShell.

Personal tooling.

Reference human-approval orchestration for REQUEST_CONFIRMATION flows — pluggable channels (webhook, Slack, Teams, email) + a replay-safe resume via adapter-core confirm().

One-command installer for the dragon375014 AI dev toolchain: idea → spec → goals → parallel execution. specmit = spec + submit + MIT (Made in Taiwan).

Read-only consistency validator: 9 deterministic drift checks adjudicated by a per-fact-class authority registry (sot-map.yaml); LLM-judged findings advisory-only; golden-fixture gated.

The Mentat engine: scope algebra, conflict derivation, cascade invalidation, and the amendment log. Pure functions, no I/O.

Local-first proof-of-build ledger for AI-assisted software. `npx vibetrace-cli` collects your local AI-agent build traces, hashes them, and publishes an inspectable public build story with a build score.

Human-rating distillation: FeedbackRecords + session turns to eval samples, the auto-distill watermark, and the daemon janitor step that keeps a <spec>-ratings registry dataset current

This package is not maintained by Eligible.com. NextStep LLC saw a need for an updated NodeJS wrapper, and created this package.

A helper library to help generate test claims data for provider-claims-api

Deterministic safe-view-only baseline bots for Duplicate Bridge.

TRIAL USE - FHIR specification for electronic submission of Pharmacy Claims for adjudication in Canada. This FHIR standard is created by CPHA, CLHIA and various Canadian stakeholder groups including vendors, Pharmacy associations, Pharmacy Benefit Manag

An adjudicator for the rules of the Diplomacy board game implemented in coffee script.

External adapters for Mentat, each behind an interface with a deterministic fallback. Nothing here requires an API key to work.

Evidence-driven adversarial reviews across AI coding agents.

VELLA — deterministic pre-execution governance substrate. SDK for producing signed decision records.

Deterministic scholarly-citation extraction, resolution, evidence capture, and audit reporting.

MCP server for medicare-coverage