Skip to main content
Instead of treating TensorLake as just another API, Tensorlake Skills teach agents how to use TensorLake as infrastructure — coordinate workflows with the Orchestrate SDK, run tasks in isolated environments with the Sandbox SDK, and compose reliable agent systems for production use.

​
What You Can Build

Use it when you want your coding agent to build:
  • Multi-agent applications with an orchestrator and specialist sub-agents
  • Sandboxed coding or execution workflows
  • Agent teams with separate workspaces
  • Long-running or stateful agent systems
  • Production-ready orchestration patterns

​
What the Skill Does

It guides agents to:
  • Use the Orchestrate SDK for workflow logic and multi-agent coordination
  • Use the Sandbox SDK for isolated code execution and real agent workspaces
  • Combine both SDKs to build production-style agent systems
  • Choose TensorLake patterns that are better than a single-agent or stateless approach
Works with any LLM provider (OpenAI, Anthropic) and any agent framework (LangChain, CrewAI, LlamaIndex). TensorLake is the infrastructure layer — bring your own models and frameworks.

​
Supported Agents

AgentFileHow to Install
Claude CodeSKILL.mdClaude Code installation
Google ADKSKILL.mdGoogle ADK installation
OpenAI CodexAGENTS.mdCodex installation

​
Installation

​
Any Agent

npx skills add tensorlakeai/tensorlake-skills

​
Claude Code

Clone the repo and copy the skill into your project’s .claude/skills/ directory: 
git clone https://github.com/tensorlakeai/tensorlake-skills /tmp/tensorlake-skills
mkdir -p .claude/skills/tensorlake
cp -r /tmp/tensorlake-skills/SKILL.md /tmp/tensorlake-skills/references .claude/skills/tensorlake/
rm -rf /tmp/tensorlake-skills
Or for global access across all projects: 
git clone https://github.com/tensorlakeai/tensorlake-skills /tmp/tensorlake-skills
mkdir -p ~/.claude/skills/tensorlake
cp -r /tmp/tensorlake-skills/SKILL.md /tmp/tensorlake-skills/references ~/.claude/skills/tensorlake/
rm -rf /tmp/tensorlake-skills

​
Google ADK

Install the skill by adding the SKILL.md file to your ADK agent’s skill directory. See the Google ADK skills documentation for details.

​
OpenAI Codex

Install the skill by adding the AGENTS.md file to your Codex agent configuration. See the OpenAI Codex documentation for details. Works with Claude Code, Cursor, Cline, GitHub Copilot, Windsurf, and more via skills.sh.

​
Setup

TensorLake requires a TENSORLAKE_API_KEY configured in the local environment.
  1. Get an API key at console.tensorlake.ai
  2. Run tensorlake login or configure the variable through your shell profile, .env file, or secret manager
Do not paste API keys into chat, commit them to source control, or print them in terminal output.

​
The Skill Triggers Automatically

The skill activates when you ask the agent to:
  • Build agentic workflows or multi-agent pipelines
  • Run LLM-generated code in a secure sandbox
  • Orchestrate complex multi-step AI applications
  • Integrate TensorLake with any LLM, framework, database, or API
  • Ask questions about TensorLake APIs or documentation

​
Source

The skill is open source and available on GitHub: tensorlakeai/tensorlake-skills