orchata-rag

What This Skill Does

The orchata-rag skill empowers your OpenClaw agent to become an expert in your organization's internal knowledge base. By integrating the Orchata RAG platform, the agent gains the ability to manage, search, and navigate large collections of documents using a sophisticated tree-based indexing structure. Unlike simple keyword search tools, Orchata parses files like PDFs, Word documents, and presentations into hierarchical sections with summaries, allowing the agent to retrieve precise, context-aware information rather than just raw text blocks. Whether you are managing technical documentation, legal case files, or project repositories, this skill transforms static documents into a dynamic, queryable asset for your AI agent.

Installation

To integrate this knowledge management powerhouse, run the following command in your terminal or OpenClaw interface:

clawhub install openclaw/skills/skills/ehudsn/orchata

Ensure that your OpenClaw environment has the necessary permissions to access the Orchata MCP server instance associated with your organizational account.

Use Cases

• Technical Documentation: Query complex technical manuals by section or page range to find specific configuration steps.
• Legal Research: Organize and retrieve case files from different "Spaces," allowing the agent to perform cross-reference analysis.
• Onboarding Assistance: Provide new hires with an AI agent that knows exactly where company policies, benefits, and architectural standards are stored.
• Project Management: Manage vast archives of project reports and stakeholder presentations, enabling quick summaries of past initiatives.

Example Prompts

1. "Look through the 'Technical Docs' space and find the section summarizing our current deployment architecture, then provide a brief explanation."
2. "List all active knowledge spaces and tell me which one would be best for researching our new privacy policy updates."
3. "Search for the latest 'Q4 Financial Review' document within the 'Finance' space and summarize the key performance indicators mentioned."

Tips & Limitations

• Status Check: Always confirm that your document status is COMPLETED before initiating a search. Documents in PENDING or PROCESSING states are invisible to query tools.
• Leverage Descriptions: Ensure your 'Spaces' are well-named and described. The smart_query function relies on these metadata fields to intelligently route user requests to the correct document container.
• Pagination: When listing spaces with large numbers of results, remember to utilize the page and pageSize parameters to avoid truncating your search results.