telnyx-rag

What This Skill Does

The telnyx-rag skill transforms your OpenClaw workspace into an intelligent, searchable database. By leveraging Telnyx's high-performance native embedding, similarity search, and inference APIs, this skill enables a complete Retrieval-Augmented Generation (RAG) pipeline locally. It automates the indexing of your workspace files—including documentation, memory logs, and project notes—chunking them intelligently to ensure that the AI provides precise, context-aware answers. Whether you are managing complex documentation or simply looking to query your personal "brain," this skill ensures that your knowledge is always at your fingertips.

Installation

To begin, ensure you have Python 3.8+ installed on your system. You can install the skill directly via the ClawHub command line utility:

clawhub install openclaw/skills/skills/dotcom-squad/telnyx-rag

Once installed, navigate to the directory and configure your credentials. Create a .env file in the skill folder containing TELNYX_API_KEY=your_key_here. Finally, run ./setup.sh to initialize your workspace bucket. Follow the naming convention openclaw-{agent-id} to maintain compatibility with standard OpenClaw organizational patterns.

Use Cases

• Knowledge Retrieval: Quickly pull specific policies or technical instructions from internal documentation without manually searching through files.
• Agent Memory Management: Allows agents to "remember" previous interactions, meeting decisions, or user-specific preferences stored in markdown files.
• Onboarding Assistance: Automate the answering of repetitive "How-to" questions for new team members by pointing the RAG pipeline at your team's existing knowledge base.
• Project Analysis: Synthesize information from multiple disparate project files to identify dependencies or missed deadlines.

Example Prompts

1. "Based on my memory logs, what were the primary constraints decided during the last project kickoff meeting?"
2. "How do I configure the new authentication headers for our internal API?"
3. "Search my project docs and summarize the deployment process for the production environment."

Tips & Limitations

• Incremental Sync: The skill is designed for efficiency; it automatically detects changes, so you don't need to re-index the entire workspace every time a small change is made.
• Orphan Cleanup: Regularly run the sync process to ensure that deleted local files are also removed from your Telnyx storage bucket.
• Model Selection: While the default model is efficient, you can use the --model flag in ./ask.py to switch to more powerful models like Llama 3.1 if you require complex reasoning.
• Context Window: Be mindful of the --num parameter; while increasing chunks provides more context, it may impact response latency and token costs.