
from tensorlake-skills176
Provides agent-aware guidance and workflows for Tensorlake sandboxes and orchestration — use when the user asks about Tensorlake APIs, sandboxing, or SDK usage.
Tensorlake SDK equips agents with authoritative guidance on using Tensorlake's stateful sandboxes and sandbox-native orchestration. The skill instructs agents to fetch live docs, verify SDK symbols, and follow guardrails for executing or orchestrating code inside isolated sandboxes (suspend/resume, snapshots, clones, PTY, browser automation, and port tunneling). It supports Python, TypeScript, and CLI flows and explains when to prefer live docs vs bundled references.
Use this skill whenever a user mentions Tensorlake, sandboxes, or asks how to run or orchestrate agent code safely. Activate when building agentic systems that need isolated execution, remote-dev setups, file transfer, or when you must verify SDK import paths and methods. The skill mandates WebFetch of the live docs index before giving actionable code.
Primarily designed for coding-capable agents (Claude Code, OpenAI Codex, Copilot-style assistants) and any agent frameworks that can run WebFetch and execute code (e.g., LangChain integrations).
This skill has not been reviewed by our automated audit pipeline yet.