Content
# YANA • Yet Another Network Agent
[](https://github.com/pdudotdev/YANA/releases/tag/v1.2.0)
[](https://github.com/pdudotdev/YANA/commits/main/)
| | |
|---|---|
| **Platforms** |         |
| **Transport** |   |
| **Integrations** |    |
## Overview
RAG-powered network QA investigation tool for multi-vendor networks.
Run your tests with any framework. When something fails, YANA investigates - it queries live devices, searches protocol specs, follows diagnostic decision trees, and tells you why it failed and what to fix.
**Supported models:**
- Haiku 4.5, Sonnet 4.6, Opus 4.6 (default, best reasoning)
**Documentation:**
- [**WORKFLOW.md**](metadata/workflow/WORKFLOW.md) - operational flow
- [**GUARDRAILS.md**](metadata/guardrails/GUARDRAILS.md) - security controls
**What's new:**
- [**CHANGELOG.md**](CHANGELOG.md) - latest version
## Tech Stack
| Technology | Role |
|-----------|------|
| Python | Core language |
| FastMCP | MCP server exposing 8 tools |
| Claude | Reasoning, context, investigation |
| LangChain | RAG pipeline (chunking, embedding, retrieval) |
| ChromaDB | Vector database for knowledge base |
| Scrapli | Multi-vendor SSH transport |
| JUnit XML | Test results format |
## Scope
| Protocol | What's Checked |
|----------|---------------|
| **OSPF** | Adjacencies, area and area types, config, LSDB |
| **Routing** | Table, route maps, prefix lists, PBR, ACLs, ECMP |
| **Interfaces** | Up/down state, expected operational status |
## Test Network Topology
**Network diagram:**
**Lab environment:**
- 16 devices defined in [**TOPOLOGY.yml**](TOPOLOGY.yml)
- 5 x Cisco IOS, 3 x Cisco IOS-XE, 4 x Arista cEOS, 2 x MikroTik CHR, 1 x Juniper JunOS, 1 x Aruba AOS-CX
- See [**lab_configs**](lab_configs/) for my test network's configuration
## Installation
**Prerequisites:** Python 3.11+
**Step 1 - Install and ingest:**
```bash
sudo apt install git make python3.12-venv -y
cd ~ && git clone https://github.com/pdudotdev/YANA
cd YANA && make setup
```
**Step 2 - Configure credentials:**
```bash
cp .env.example .env
# Edit .env with your device credentials
```
**Step 3 - Authenticate with Claude:**
```bash
claude auth login
```
**Step 4 - Register the MCP server:**
```bash
claude mcp add yana -s user -- /home/<user>/YANA/yana/bin/python /home/<user>/YANA/server/MCPServer.py
```
## MCP Tools
| Tool | Description |
|------|-------------|
| `get_status` | Report active data sources (inventory, intent, KB) |
| `list_devices` | List inventory devices, optionally filtered by CLI style |
| `get_ospf` | Query OSPF state (neighbors, LSDB, process config) on a live device |
| `get_interfaces` | Query interface up/down status on a live device |
| `get_routing` | Query routing table, route maps, prefix lists, PBR, and ACLs on a live device |
| `query_intent` | Retrieve network design intent from local JSON |
| `traceroute` | Trace the forwarding path from a device to a destination IP |
| `search_knowledge_base` | Search network knowledge base (RFCs, vendor guides) with protocol/vendor/topic filters |
## Customization
YANA is designed to work with your own test topology. Bring your own:
| What | Where | Format |
|------|-------|--------|
| **Device inventory** | `data/NETWORK.json` | JSON dict: device name -> host, platform, cli_style |
| **Design intent** | `data/INTENT.json` | JSON: router roles, OSPF areas, links, neighbors |
| **Protocol docs** | `docs/*.md` | Protocol data (re-ingest with `make ingest`) |
| **Test results** | `results/*.xml` | JUnit XML from any test runner |
| **Credentials** | `.env` | `ROUTER_USERNAME` and `ROUTER_PASSWORD` |
## QA Workflow
1. **Run your tests** with any framework (pytest, pyATS, Ansible, etc.)
2. **Place JUnit XML results** in `results/`
3. **Investigate failures** with the `/qa` skill:
```
claude
> /qa
```
The `/qa` skill loads the latest JUnit XML, lists all failures, and investigates each one using live device queries, network intent, and RFC context.
## Knowledge Base
Protocol documentation lives in `docs/` as Markdown files. Each file is tagged with `vendor`, `topic`, and `protocol` metadata during ingestion.
To update after editing docs:
```bash
make ingest
```
**RAG optimizations in place:**
- `protocol` metadata field - filters search by protocol (ospf, bgp, eigrp), eliminating cross-protocol noise
- Contextual chunk headers - source and protocol prepended to each chunk for better embedding quality
- Compound filtering - combine vendor, topic, and protocol filters in a single query
See [**OPTIMIZATIONS.md**](metadata/scalability/OPTIMIZATIONS.md) for the full optimization roadmap.
## Project Structure
```
YANA/
├── server/
│ └── MCPServer.py # FastMCP server (8 tools)
├── tools/ # MCP tool implementations
│ ├── ospf.py # get_ospf
│ ├── operational.py # get_interfaces, traceroute
│ ├── routing.py # get_routing
│ ├── intent.py # query_intent
│ ├── status.py # get_status
│ ├── inventory_tool.py # list_devices
│ └── rag.py # search_knowledge_base
├── core/ # Infrastructure
│ ├── inventory.py # Device dict (from data/NETWORK.json)
│ └── settings.py # Credentials + SSH config
├── data/ # Data files
│ ├── NETWORK.json # Device inventory
│ ├── INTENT.json # Network design intent
│ └── chroma/ # ChromaDB vector store (generated)
├── transport/ # SSH execution (Scrapli)
│ ├── __init__.py # Command dispatcher + semaphore
│ └── ssh.py # Multi-vendor SSH executor
├── platforms/ # Vendor CLI command mapping
│ ├── platform_map.py # OSPF, routing, interface + traceroute commands (6 vendors)
│ └── definitions/ # Custom Scrapli definitions
├── input_models/
│ └── models.py # Pydantic validation (OspfQuery, KBQuery, etc.)
├── docs/ # Knowledge base (RFCs + vendor guides)
├── results/ # QA test results (JUnit XML)
├── skills/
│ ├── ospf/ # OSPF diagnostic decision tree
│ └── routing/ # Routing diagnostic decision tree
├── testing/
│ ├── automated/ # Unit + integration tests
│ ├── live/ # Live lab tests
│ └── run_tests.sh # Test runner (--live for lab tests)
├── metadata/
│ ├── guardrails/ # Security controls documentation
│ ├── scalability/ # RAG optimization roadmap
│ ├── topology/ # Test network diagram
│ └── workflow/ # RAG pipeline walkthrough
├── ingest.py # RAG ingestion pipeline
├── Makefile # Setup automation (make setup)
├── CLAUDE.md # Investigation workflow + tool reference
├── TOPOLOGY.yml # Containerlab topology definition
├── .env.example # Credential template
├── CHANGELOG.md # Version history
├── LICENSE
├── requirements.txt
└── README.md
```
## Planned Upgrades
- EIGRP and BGP support (tools, skills, knowledge base docs)
## Disclaimer
You are responsible for defining your own network inventory and design intent, building your test environment, and meeting the necessary prerequisites (Python 3.11+, Claude CLI/API, network device access).
## License
Licensed under [**GNUv3.0**](LICENSE).
## Collaborations
Interested in collaborating?
- **Email:** [**hello@ainoc.dev**](mailto:hello@ainoc.dev)
- **LinkedIn:** [**LinkedIn**](https://www.linkedin.com/in/tmihaicatalin/)
