https://github.com/satunix/aatis-skms
Private and Local - Smart Knowledge Management System - SKMS
https://github.com/satunix/aatis-skms
autogen localai ollama python
Last synced: 3 months ago
JSON representation
Private and Local - Smart Knowledge Management System - SKMS
- Host: GitHub
- URL: https://github.com/satunix/aatis-skms
- Owner: SATUNIX
- Created: 2025-06-06T03:49:10.000Z (4 months ago)
- Default Branch: main
- Last Pushed: 2025-07-08T10:57:23.000Z (3 months ago)
- Last Synced: 2025-07-08T11:49:37.939Z (3 months ago)
- Topics: autogen, localai, ollama, python
- Language: Python
- Homepage:
- Size: 84 KB
- Stars: 0
- Watchers: 0
- Forks: 0
- Open Issues: 2
-
Metadata Files:
- Readme: README.md
Awesome Lists containing this project
README
# SKMS AATIS Module
Agentic Assistance for Testing Information Systems
Smart Knowledge Management System 0.0.1
> Fully Local & Privacy Friendly, (only outbound is websearches via Bing Web Search API)
See issues page for current known issues.
A single Python-based CLI tool to:
1. **generate**: Search the web and RAG store, summarize results into a timestamped markdown report.
2. **ask**: Run a RAG-powered Question & Answer session over your previously saved reports.
3. **refresh**: Ingest all markdown (`.md`) reports into a FAISS vector store for future queries.
This tool utilizes AutoGen 0.6.1, the Bing Web Search API (search engine), Ollama (language model), and FAISS (vector database).
---
```mermaid
flowchart TD
main[main.py CLI] -->|generate| gen[generate_agent.py]
main -->|ask| ask[qa_agent.py]
main -->|refresh| refresh[ingest_reports.py]
gen --> search[BingSearchTool]
gen --> fetch[FetchWebpageTool]
gen --> rag1[RagStore]
gen --> model1[OllamaChatCompletionClient]
ask --> rag2[RagStore]
ask --> model2[OllamaChatCompletionClient]
refresh --> rag3[RagStore]
```
## Prerequisites
- Python 3.8 or newer
- Git
- **Bing Search API Key**: Set the `BING_SEARCH_API_KEY` environment variable to your Bing Search API key.
- **Bing Search Endpoint (optional)**: Set the `BING_SEARCH_ENDPOINT` environment variable (default: `https://api.bing.microsoft.com/v7.0/search`).
- Running Ollama server (default URL: `http://localhost:11434`)
---
## Quickstart Guide
### Step 1: Clone the Repository
```bash
git clone https://your.repo.url/project.git
cd project
```
### Step 2: Setup Virtual Environment
Linux/macOS:
```bash
python3 -m venv .venv
source .venv/bin/activate
```
Windows:
```bash
python -m venv .venv
.venv\Scriptsctivate
```
### Step 3: Install Dependencies
```bash
pip install --upgrade pip
pip install autogen-agentchat==0.6.1 "autogen-ext[ollama]" requests beautifulsoup4 sentence-transformers faiss-cpu
```
### Step 4: Create Initialization Files
```bash
touch tools/__init__.py agents/__init__.py rag/__init__.py
```
### Step 5: Configure Ollama Client
Edit the file `config/ollama_config.json`:
```json
{
"model": "llama3",
"base_url": "http://localhost:11434"
}
```
---
## Usage Instructions
### Generate a New Report
This command searches the web, summarizes findings using Ollama, and saves the output as a timestamped markdown file:
```bash
python main.py generate "quantum computing breakthroughs"
```
### Refresh RAG Database
Ingest all existing markdown reports into the FAISS database:
```bash
python main.py refresh
```
### Ask a Question
Use this to query your RAG database and model knowledge:
```bash
python main.py ask "What did we learn about quantum entanglement?"
```
---
## Configuration Options
- **Bing Search API**: Provide your API key via the `BING_SEARCH_API_KEY` environment variable and optionally override the endpoint with `BING_SEARCH_ENDPOINT`.
- **Ollama model settings**: Adjust in `config/ollama_config.json`.
- **Embedding model**: Set in `rag/rag_store.py` (default is `all-MiniLM-L6-v2`).
---
## Troubleshooting Common Issues
- Connection errors to the Bing Search API or Ollama: Ensure your API key is valid, environment variables are set, and endpoints are correct.
- No reports found during refresh: Confirm that you have run the `generate` command at least once.
- FAISS errors: Check file permissions or delete and regenerate the `rag/index.faiss` file.
---
You are now ready to use the Unified RAG & Research CLI tool.