Client SDK for dmstfy-rag: grounded RAG queries with citations, groundedness, abstention and SSE streaming

Playwright assertion matchers for testing LLM outputs — groundedness, PII, sentiment, schema, and fuzzy matching

Agent trust guardrails — PII detection/redaction (Luhn-checked), prompt-injection detection, groundedness, and a tool-output safety filter. Pure, deterministic, dependency-free.

Compliance-native evaluation library for LLM applications

Multi-agent simulation harness that stress-tests conversational AI agents with adversarial synthetic users, then grades every transcript with auditable decision-tree judges.

Glass Box Framework MCP server — runtime constitutional verification for AI answers, with claim-level reasoning chains, formal ECS scoring, 7-angle red team (Glassbox Court), and deterministic audit logs.

Knowledge ingestion, local memory, RAG, HyDE, retrieval, context packages, and vault materialization for NetToolsKit.

Zero-config Node.js wrapper for M365 Copilot Agent Evaluations CLI (Python-based Azure AI Evaluation SDK)

TypeScript client for the Babav Knowledge Core API (Deno + Node 18+, zero deps). Includes the babav.visual grammar TYPES at the ./visual subpath (types only; all visual rendering is server-side).

ISI MCP Server — Persistent AI Identity Infrastructure powered by CCSP. Session lifecycle, contradiction detection, grounding verification, behavioral signature computation, and identity flywheel.

Testing utilities, mock providers, fixtures, and eval framework for ElsiumAI

prism-verify — runtime LLM verifier (family-different, reasoning-stripped, multi-lens, signed Ed25519 receipts). Zero-prerequisite npx install via a verified binary launcher.

MCP server for ThinkHive Explainer API v3 - Run-centric architecture with facts vs inferences, deterministic linking, calibrated predictions, and AI agent observability

350 built-in guards for LLM safety: 26 PII regions, prompt injection, toxicity, bias, agent safety, hallucination, content safety, compliance, format validation, Korean ISMS-P/PIPA, GDPR, EU AI Act

Production caching seam for AI workloads (keyless/offline memory reference + Redis adapter): exact-match cache, verified semantic cache, embedding cache, singleflight request coalescing, negative cache, and epoch-based invalidation. Makes the gateway + RA

AI Agent Observability & Security — MCP server for tracing reasoning chains, tracking costs, detecting hallucinations, and monitoring agent behavior

We help GenAI teams maintain high-accuracy for their Models in production.

Scaffold a production-grade AI feature spec before you write a line of code — spec, eval criteria, golden test cases, and an ADR starter

MCP server for grounding / hallucination checking: verify a claim against evidence (Supported/Refuted/Not Enough Evidence) — for AI agents.

Deterministic + semantic verification of quote-grounded LLM citations (EVI1 protocol): fuzzy verbatim-quote matching and an LLM entailment judge for transparent, per-claim hallucination detection.