Manage LLM agent lifecycles with YAML/Markdown configurations and tool integrations.
npx skills add https://github.com/treasure-data/td-skills --skill agent-treasure-data
Please help me install this Agent Skill. Skill: agent Source: https://github.com/treasure-data/td-skills/tree/main/tdx-skills/agent Command: npx skills add https://github.com/treasure-data/td-skills --skill agent-treasure-data
This Skill helps you build and manage LLM agents using Treasure Data tooling. It demonstrates a pull/push workflow and YAML/Markdown configurations to streamline agent development.
Use the agent skill to pull a project: tdx agent pull "My LLM Project" and then push changes: tdx agent push.
High-intent search queries and answers about installing and using this skill.
Automate LLM agent workflows by defining agent.yml with system prompts, starter messages, and tool references. Use tdx agent pull to fetch your project, edit the YAML structure with knowledge_base, web_search, and image_gen tool integrations, then push changes back with tdx agent push for reproducible deployment.
Agent lifecycle management uses a pull/push workflow: retrieve your project with tdx agent pull, modify agent configurations and knowledge bases locally, then synchronize changes using tdx agent push. This pattern ensures version control and reproducible agent deployment across your TD environment.
Configure multiple tools in your agent.yml by declaring knowledge_base, agent, web_search, and image_gen as tool components with @ref cross-references. The Skill structures these integrations within your TD folder layout, enabling multi-tool capabilities in a single agent configuration.
Yes, agent configuration supports both Markdown and YAML formats (.md and .yml files) within your project structure. Use Markdown for knowledge bases and documentation, YAML for agent.yml structure and tool definitions, with @ref syntax to link and reference configurations across files.
End-to-end agent deployment requires setting up TD folder layout with agent.yml, defining system prompts and starter messages, organizing knowledge bases in .md/.yml, configuring tool integrations, and using pull/push operations. The Skill guides this complete workflow from initial setup through reproducible deployment.
The Skill demonstrates LLM agent development specifically within Treasure Data's tooling and YAML/Markdown configuration patterns. Familiarity with YAML syntax and basic agent concepts helps, but the workflow is designed to streamline agent construction for TD users without requiring external agent framework expertise.
Snowflake-LabsA concise summary of your roleplay character: name,</name> personality, and/or behavior, formatted as a single string sentence. Typically不超过50字，此处省略
richardnguyen0715Triggers Hermes agent runs via webhooks from external services like GitHub and Stripe.
SteveFeldmanSearch the web, evaluate source credibility, and synthesize cited findings.
kagent-devBuild, deploy, and manage AI agents natively on Kubernetes.
