create-llama
CLI scaffolder for full-stack LlamaIndex RAG and agentic apps — Next.js, FastAPI, or Express in one command.
create-llama
create-llama is an interactive CLI that bootstraps production-ready LlamaIndex applications in seconds. It generates a fully wired full-stack project — your choice of Next.js, Python FastAPI, or Express backend — with document ingestion, vector storage, chat UI, streaming, and optional agentic tools pre-configured.
Key Features
- Full-stack templates: Next.js + Python FastAPI, Next.js + Express, or Python API-only
- One-command setup: interactive prompts select model provider (OpenAI, Anthropic, Groq, Ollama), vector store (Pinecone, Chroma, Qdrant, in-memory), and data sources
- Streaming chat UI: complete React chat interface with streaming tokens, file upload, and multi-turn history
- Document pipelines: PDF, Markdown, HTML, CSV ingestion with chunking + embedding baked in
- Agentic mode: toggle agent mode to add tool-calling loops with DuckDuckGo search, code interpreter, and custom tools
- Production-ready: Docker Compose, Vercel deploy button, and
.envmanagement all included
Quick Start
npx create-llama@latest
# Interactive wizard: pick stack, model, vector store
# Then:
cd my-llama-app && npm run dev
npx ai-supply add create-llama-fullstack-scaffold
Curated mirror of the open-source create-llama (MIT). Get it from the source.
Compromise signals — malicious or tampered code (leaked secrets, backdoors, a dropped executable) — reduce the score, and known dependency CVEs carry a bounded penalty (they warrant review but never QUARANTINE — update the dependency to clear). Other dangerous-by-capability traits are risk surface, expected for some capabilities. Every finding is mapped to its OWASP control below.
Findings mapped to the OWASP Top 10 for LLM Applications (2025) and the OWASP Machine Learning Security Top 10. Expand any flagged control for the exact findings — compromise reduces the score; expected/risk-surface do not, except a known CVE, which carries a small bounded penalty (high/critical → Review).
The same gate an agent runs before installing (POST /api/v1/trust/create-llama-fullstack-scaffold/check). Click a policy:
Consume create-llama programmatically. Authenticate with an API key or session — see Authorize an agent.
# Agents: CHECK BEFORE YOU INSTALL (no auth) — score, grade, level, capability manifest
curl https://ai-supply.store/api/v1/trust/create-llama-fullstack-scaffold
# Gate against your org policy (returns { pass, violations })
curl -X POST https://ai-supply.store/api/v1/trust/create-llama-fullstack-scaffold/check \
-H "Content-Type: application/json" \
-d '{"minGrade":"B","denyPermissions":["shell"],"denyUnknownEgress":true}'
# CLI
npx ai-supply add create-llama-fullstack-scaffold
# REST (install → download)
curl -X POST https://ai-supply.store/api/v1/listings/create-llama-fullstack-scaffold/install \
-H "Authorization: Bearer $AIM_KEY"
# MCP tool
install_listing({ "slug": "create-llama-fullstack-scaffold" })OpenAPI spec →Curated mirror — latest upstream source. See the repository for tagged releases.