feat: adding LlamaStack integration with chat, embeddings, and safety

[omaryashraf5]

Description: This PR introduces a comprehensive LangChain integration for Llama Stack, providing full-featured chat completion, text embeddings, and safety checking capabilities. The integration supports multiple providers (Ollama, OpenAI, Together AI, Fireworks, etc.) and includes advanced shield management for content moderation.

Key features:

• Chat Completion: Full LangChain-compatible chat models with streaming support
• Text Embeddings: Vector embeddings for semantic search and RAG applications
• Safety Checking: Content moderation using Llama Guard and other safety shields with registration capability
• Shield Management: Register, and clean up safety shields with automatic provider detection
• Multi-Provider Support: Works with Ollama, OpenAI, Together AI, Fireworks, and more
• Comprehensive Examples: Complete usage examples including basic usage, advanced features, and shield management
• Utility Functions: Model discovery, connection testing, and environment validation

The integration follows LangChain patterns and includes proper error handling, logging, and comprehensive documentation.

Dependencies:

• llama-stack-client>=0.0.40 (optional, for Llama Stack integration)
• httpx>=0.25.0
• pydantic>=1.10.0
• numpy and scikit-learn (optional, for similarity search)

Additional Notes:

• All tests pass locally with make format, make lint, and make test
• Integration includes comprehensive unit tests that don't rely on network access
• Example notebooks and documentation are included in the examples/ directory
• Optional dependencies are properly imported within functions
• Changes are backwards compatible
• No modifications to existing packages, this is a self-contained new integration

[vercel]

The latest updates on your projects. Learn more about Vercel for GitHub.

Project	Deployment	Preview	Comments	Updated (UTC)
langchain	Ready	Preview	Comment	Aug 27, 2025 11:58pm

[codspeed-hq]

CodSpeed WallTime Performance Report

Merging #32706 will not alter performance

_{Comparing omaryashraf5:langchain_llamastack (0d72193) with master (7f9b077)}

⚠️ Unknown Walltime execution environment detected

Using the Walltime instrument on standard Hosted Runners will lead to inconsistent data.

For the most accurate results, we recommend using CodSpeed Macro Runners: bare-metal machines fine-tuned for performance measurement consistency.

Summary

✅ 13 untouched benchmarks

[codspeed-hq]

CodSpeed Instrumentation Performance Report

Merging #32706 will not alter performance

_{Comparing omaryashraf5:langchain_llamastack (0d72193) with master (7f9b077)}

Summary

✅ 14 untouched benchmarks