# AI Agent Operations

> AI model configuration, LLM memoization, agentic RAG search, and vector library maintenance.

- Skill: `gitwalter/ai-agent-operations` (Agent Skill)
- Install (CLI): `npx skillmds@latest add gitwalter/ai-agent-operations`
- Raw SKILL.md: https://api.skillmd.com/api/skills/gitwalter/ai-agent-operations/raw
- Safety review: pending
- Works with: Claude Code, Claude.ai, OpenAI Codex
- Category: AI & ML
- Author: gitwalter (https://skillmd.com/u/gitwalter)
- Updated: 2026-09-10
- Page: https://skillmd.com/skills/gitwalter/ai-agent-operations

---


# AI Agent & RAG System Operations

This skill covers the configuration and operational execution of Large Language Models (LLM), response memoization layers, agentic RAG searches, and vector database document library maintenance.

## When to Use
Use this skill when configuring LLM endpoints, setting up response memoization caches, running CLI RAG searches, rebuilding library Tables of Contents (TOC), or repairing vector indices.

## Prerequisites
- Conda environment initialized.
- Qdrant Vector database or OpenAI API credentials set in environmental variables.

## Process

Follow these procedures to query RAG models and index agent knowledge.

### Executing RAG Search via CLI
Query the vector store directly:
```bash
conda run -p D:\Anaconda\envs\cursor-factory python scripts/ai/rag/rag_cli.py --query "What is the 5-layer architecture?"
```

### Rebuilding Library TOCs
Regenerate document tables of contents for optimal chunk indexing:
```bash
conda run -p D:\Anaconda\envs\cursor-factory python scripts/ai/rag/rebuild_tocs.py
```

## Best Practices
- **Optimize Prompt Tokens**: Leverage the memoization layer to cache repetitive API requests.
- **Library Audits**: Routinely run `repair_library.py` to fix missing document metadata.