Trilogy AI COE Logo Trilogy AI Center of Excellence

Resources

Access our curated collection of tools, datasets, and prompt libraries to accelerate your AI research and development.

Featured Resource

Capture how engineering agents work, connect behavioral assessments to native trajectory evidence, and compare repeated trials across harnesses. EBO brings workspace outcomes, LLM-as-judge evaluation, and human review into an inspectable record of each task.

  • Retains native messages, tool activity, usage, and workspace changes with explicit capture-quality checks.
  • Grounds behavioral judgments in cited evidence, with optional human review and calibration.
  • Compares matched trials in the Behavior Atlas and produces reviewable reports.
GitHub Repo Quickstart
Behavioral assessments and measured timing
Tools
Side-by-side examples of Behavior Atlas assessments for DeepSeek and Pi, and task duration across Pi, DeepSeek Harness, and Codex using Grok 4.7 and Grok 4.6.
Behavioral assessments and measured timing

Capture how engineering agents work, connect behavioral assessments to native trajectory evidence, and compare repeated trials across harnesses. EBO brings workspace outcomes, LLM-as-judge evaluation, and human review into an inspectable record of each task.

  • Retains native messages, tool activity, usage, and workspace changes with explicit capture-quality checks.
  • Grounds behavioral judgments in cited evidence, with optional human review and calibration.
  • Compares matched trials in the Behavior Atlas and produces reviewable reports.
Agent EvaluationObservability
GitHub Repo Quickstart
Reference CybersecurityAI Security

A practical guide to baseline security controls, AI-assisted testing, and incident response. Includes security harness workflows, local exercises, operational checklists, and a 30/60/90-day rollout for builders, operators, and security professionals.

Read Playbook GitHub Repo

The AI Cybersecurity Playbook

Reference
The AI Cybersecurity Playbook

A practical guide to baseline security controls, AI-assisted testing, and incident response. Includes security harness workflows, local exercises, operational checklists, and a 30/60/90-day rollout for builders, operators, and security professionals.

CybersecurityAI Security
Read Playbook GitHub Repo
Tools RustOrchestrationAgentic Engineering

A control plane for supervised agent work that connects issue tracking, isolated workspaces, execution, review, and recovery into one inspectable loop.

Visit Website GitHub Repo Read Article

OpenSymphony

Tools
OpenSymphony

A control plane for supervised agent work that connects issue tracking, isolated workspaces, execution, review, and recovery into one inspectable loop.

RustOrchestrationAgentic Engineering
Visit Website GitHub Repo Read Article
Tools PythonVideo Generation

Automated long-form text-to-video generation pipeline that overcomes limitations in current text-to-video generation models through advanced techniques like video segment chaining and keyframe-based generation.

GitHub Repo

TTV Pipeline

Tools
TTV Pipeline

Automated long-form text-to-video generation pipeline that overcomes limitations in current text-to-video generation models through advanced techniques like video segment chaining and keyframe-based generation.

PythonVideo Generation
GitHub Repo
Tools MediaToolingSkills

Reusable tooling and agent skills for media production workflows, including subtitle pipelines, corpus ingest, and rough-cut assembly.

GitHub Repo

Media-Tooling Toolkit

Tools
Media-Tooling Toolkit

Reusable tooling and agent skills for media production workflows, including subtitle pipelines, corpus ingest, and rough-cut assembly.

MediaToolingSkills
GitHub Repo
Reference DesignStandardsSkills

A reference site for AI-assisted design standards, reusable skills, and interface guidance used to keep agent-produced work coherent.

Visit Website

Design Standards and Skills

Reference
Design Standards and Skills

A reference site for AI-assisted design standards, reusable skills, and interface guidance used to keep agent-produced work coherent.

DesignStandardsSkills
Visit Website
Tools OpenClawMulti-TenantOperations

A multi-tenant abstraction layer for running isolated OpenClaw installations on a shared host without forking the upstream container image.

GitHub Repo Watch Video

OpenClaw Tenant Wrapper

Tools
OpenClaw Tenant Wrapper

A multi-tenant abstraction layer for running isolated OpenClaw installations on a shared host without forking the upstream container image.

OpenClawMulti-TenantOperations
GitHub Repo Watch Video
Protocol FederationAgent Infrastructure

A federation daemon for AI agent gateways, with signed peer-to-peer messaging, approval flows, and relay into local agent systems.

GitHub Repo Read Article

Open Gateway Protocol

Protocol
Open Gateway Protocol

A federation daemon for AI agent gateways, with signed peer-to-peer messaging, approval flows, and relay into local agent systems.

FederationAgent Infrastructure
GitHub Repo Read Article
Tools PythonPyTorch

A robust framework implementing reference-free LLM evaluation metrics, including G-Eval, for reliable model assessment without needing human references.

GitHub Repo View Demo

Reference-Free LLM Evaluation

Tools
Reference-Free LLM Evaluation

A robust framework implementing reference-free LLM evaluation metrics, including G-Eval, for reliable model assessment without needing human references.

PythonPyTorch
GitHub Repo View Demo
Reference PythonGraph DatabaseRAG

A toolkit for implementing graph-based retrieval-augmented generation to enhance context management and relevance in complex AI applications.

GitHub Repo View Demo Read Article

GraphRAG Proof of Concept

Reference
GraphRAG Proof of Concept

A toolkit for implementing graph-based retrieval-augmented generation to enhance context management and relevance in complex AI applications.

PythonGraph DatabaseRAG
GitHub Repo View Demo Read Article