SmythOS Runtime Environment (SRE)
Open-source runtime, SDK and CLI for building, running and managing production AI agents across local, cloud and edge environments.
At a Glance
MIT-licensed SmythOS Runtime Environment (SRE) kernel, SDK and CLI for running and building AI agents with code; self-hosted.
Engagement
Available On
Alternatives
Listed Nov 2026
About SmythOS Runtime Environment (SRE)
SmythOS Runtime Environment (SRE) is an open-source runtime from SmythOS for building and running AI agents. The repository bundles the SRE kernel, a TypeScript SDK and a CLI, and is released under the MIT license. It exposes a unified API over LLMs, vector databases, storage and caches so agent logic can stay the same as infrastructure changes.
What It Is
SRE is an agent runtime and SDK that the project describes as an operating system layer for agentic AI. Developers write agents in code with the SDK, or load and run .smyth agent files created in the SmythOS visual builder. The runtime manages the agent lifecycle, resources and orchestration.
How the Resource Abstraction Works
Every resource type (storage, LLM, VectorDB, cache, vault) is accessed through a consistent interface, so providers can be swapped without changing business logic. The documented connectors are:
- Storage: Local, S3, Google Cloud, Azure
- LLM: OpenAI, Anthropic, Google AI, AWS Bedrock, Groq, Perplexity
- VectorDB: Pinecone, Milvus, RAMVec
- Cache: RAM, Redis
- Vault: JSON File, AWS Secrets Manager, HashiCorp
Security is handled through a Candidate/ACL system, which requires authorization for each resource operation.
Building Agents
The SDK supports prompt, streaming and chat modes. It also supports custom skills that combine an LLM, a vector database and storage. The README includes an article-writer agent example that does this. More than 40 components can be invoked in code or through the workflow file, including GenAILLM, APICall, WebSearch, WebScrape, DataSourceIndexer, ForEach and ServerlessCode.
Setup Path
Install the CLI with npm i -g @smythos/cli and run sre create to scaffold a project. Alternatively, add @smythos/sdk to an existing project with npm. Setting LOG_LEVEL to debug helps diagnose problems.
Community Discussions
Be the first to start a conversation about SmythOS Runtime Environment (SRE)
Share your experience with SmythOS Runtime Environment (SRE), ask questions, or help others learn from your insights.
Pricing
Open Source (SRE, SDK, CLI)
MIT-licensed SmythOS Runtime Environment (SRE) kernel, SDK and CLI for running and building AI agents with code; self-hosted.
- Smyth Runtime Environment (SRE) kernel, SDK and CLI
- Licensed under the MIT License
- 40+ production-ready components
- Pluggable connectors for storage, LLM, VectorDB, cache and vault
- Built-in Candidate/ACL security system
Capabilities
Key Features
- Unified resource abstraction across LLMs, vector databases, storage and cache
- Candidate/ACL-based secure resource access
- 40+ production-ready components
- TypeScript SDK with prompt, streaming and chat modes
- Load and run .smyth agent files
- CLI for project scaffolding
- Modular pluggable connectors
- Runs locally, in cloud, edge or hybrid environments
Integrations
Demo Video

