nanohype

A tool-agnostic template catalog for AI-focused projects. Each template is a self-contained, scaffoldable starting point — consumable by any scaffolding tool that reads the template contract.

New here?

• Building an AI client (Bedrock agent, Claude session, custom orchestrator) and want to ship software on this stack? Start with the Platform Reference and the per-repo AGENTS.md.
• Scaffolding a project locally? Start with the Quick Start guide — find the right template, scaffold it, and run it in under 2 minutes.

Templates

Templates span four engineering categories plus composable modules. Beyond engineering, the catalog also carries cross-functional templates and agent briefs for design, QA, product, marketing, sales, operations, and customer success — see Cross-Functional Templates below.

AI Systems

Template	Description	Tags
agentic-loop	Autonomous agent with tool registry, memory, provider abstraction, eval harness	typescript agent llm
rag-pipeline	TypeScript RAG pipeline with chunking, embedding, vector store, retrieval, generation	typescript rag embeddings
mcp-server-ts	TypeScript MCP server with tool/resource registration and Zod validation	mcp typescript tools
mcp-server-python	Python MCP server with tool/resource registration	mcp python tools
eval-harness	Standalone LLM evaluation framework with YAML suites, assertions, and reporters	typescript eval testing
prompt-library	Versioned prompt management with YAML frontmatter and optional TypeScript SDK	prompts yaml versioning
a2a-agent	Agent-to-Agent protocol peer with skill registry, transport abstraction, and discovery	typescript a2a agents
guardrails	Input/output safety filters — prompt injection, PII, content policy, token limits	typescript safety filters
multimodal-pipeline	Process images, audio, and video with AI models	typescript multimodal ai
fine-tune-pipeline	LLM fine-tuning pipeline with dataset preparation and training	typescript fine-tuning llm
data-pipeline	Data ingestion and transformation pipeline	typescript data pipeline
agent-orchestrator	Multi-agent orchestration with delegation and coordination	typescript agents orchestrator
ci-eval	CI-integrated LLM evaluation with GitHub Actions	typescript eval ci

Applications