A professional-grade interface for managing RAGFlow datasets, document ingestion, and high-performance chunk retrieval.
Install & Download
1. ClawHub CLI
The fastest way to install a skill directly from the registry.
npx clawhub@latest install rrragflow-skill
2. Manual Installation
Copy the skill folder to one of these locations
~/.openclaw/skills/ <project>/skills/ Priority: Workspace > Local > Bundled
3. Prompt Installation
Copy this prompt to OpenClaw to install it automatically.
Help me install rrragflow-skill using Clawhub. If Clawhub is not installed, install it first (npm i -g clawhub).
Prefer to download?
Get the raw skill files in a ZIP archive.
What is RAGFlow Dataset Ingest?
The RAGFlow Dataset Ingest skill is a specialized technical module designed for AI agents to interact seamlessly with RAGFlow environments. It provides a robust set of scripts to handle the entire lifecycle of a Retrieval-Augmented Generation (RAG) knowledge base, from creating datasets and uploading documents to managing complex parsing tasks. As a key component of the Openclaw Skills ecosystem, it bridges the gap between raw data and searchable AI-ready intelligence.
This skill ensures data integrity by enforcing strict ID resolution and providing JSON-formatted outputs that prevent information loss during relay. By utilizing this skill, developers can automate the management of their AI's long-term memory, ensuring that documents are correctly ingested, parsed, and indexed for high-accuracy retrieval during search operations.
RAGFlow Dataset Ingest Use Cases
- Creating, updating, or deleting RAGFlow datasets via CLI or automated agent workflows.
- Bulk uploading PDF or text documents to specific knowledge bases for AI training.
- Monitoring and managing the parsing status of uploaded documents to ensure high-quality chunking.
- Performing granular retrieval tests and searches across multiple datasets to validate RAG accuracy.
- Synchronizing local documentation with remote RAGFlow instances to keep AI agents up to date.
How RAGFlow Dataset Ingest Works
- The agent resolves the specific target dataset or document IDs using the list or info commands to ensure precision.
- Commands are executed through a suite of specialized Python scripts located in the scripts directory, utilizing the RAGFlow API.
- The skill requests JSON responses from the API to maintain data fidelity and prevent guessing of missing details.
- Parsing is handled as an asynchronous process; the agent initiates the parse and uses separate status checks to track progress.
- Search and retrieval operations allow for filtering by multiple dataset and document IDs to provide highly relevant context to the LLM.
RAGFlow Dataset Ingest Setup
To integrate this into your Openclaw Skills, ensure you have the required environment variables and dependencies configured:
# Install required system dependencies
# Ensure python3 is available in your environment
# Set your RAGFlow API credentials
export RAGFLOW_API_URL="your_ragflow_endpoint"
export RAGFLOW_API_KEY="your_api_key"
# Test the installation by listing datasets
python3 scripts/datasets.py list --json
RAGFlow Dataset Ingest Data Schema & Taxonomy
The skill organizes data using the standard RAGFlow metadata taxonomy, accessible via the --json flag for exact field mapping:
| Object | Key Metadata Fields |
|---|---|
| Dataset | id, name, description, document_count |
| Document | id, dataset_id, name, parse_status, progress_msg |
| Search | chunk_id, content_with_weight, relevance_score |
| Models | model_name, model_type, provider |
RAGFlow Dataset Ingest Advanced Features
- Multi-dataset search: Query across several knowledge bases simultaneously using comma-separated IDs.
- Granular Parse Tracking: Monitor parsing progress at the dataset or specific document level with error surfacing for failed tasks.
- Retrieval Testing: Dedicated mode for debugging chunking and ranking performance on single datasets.
- Safety Guardrails: Integrated confirmation steps for destructive actions and strict ID verification to prevent accidental data loss.
- Model Inspection: Programmatic access to the list of configured models within the RAGFlow instance.
SKILL.md
Loading
Related Openclaw Skills
Web Search Skill
A high-performance, paid web search proxy that provides AI agents with structured search results, rich metadata, and custom ranking capabilities.
FirstData
FirstData acts as an authoritative data navigator that connects AI agents and users to verified primary data sources, official APIs, and government portals.
Amazon Sales Estimator
A powerful analytics tool to estimate monthly Amazon sales volume and revenue across 12 marketplaces using BSR, ASIN lookups, or keyword market analysis.
Email Outreach Templates
A comprehensive library of 50+ professional business outreach templates designed to optimize every stage of the sales and partnership funnel.
Social Brand Voice
A comprehensive brand voice engine that generates tone guides, vocabulary rules, and platform-specific writing styles for consistent social media presence.
Social Engagement Audit
A comprehensive diagnostic tool that analyzes social media engagement rates and provides a structured 30-day action plan to improve reach and interaction.
Featured*
