Open Source AI Visibility Blueprint
An MIT-licensed reference engine for measuring brand presence in AI answers. Inspect the approach, run it locally, and build your own stack. Hosted Genwolf is a separate, further developed product — not this repo as SaaS.
Why this blueprint exists
AI visibility tools shouldn't be black boxes. If you can't inspect the evaluator, you can't trust the metric.
Measuring brand presence in AI-generated answers requires:
- Clear evaluation logic
- Repeatable prompt execution
- Transparent detection rules
Genwolf OSS publishes an inspectable, extensible approach to that measurement. It is a reference implementation — not a source dump of the hosted product.
Every team has a different tech stack, different internal tools, and different workflows. A one-size-fits-all SaaS can't always account for that.
With the OSS repo, you can build your own integrations on top — connect it to your CI pipelines, internal dashboards, Slack alerts, or custom reporting. Treat it as an engine and shape everything around it to fit your workflow.
You also get full freedom over the model layer. Swap in local LLMs, private endpoints, or any provider you want. It's your infrastructure, your models, your rules. The only costs are your own infrastructure and the LLM API usage you choose to run.
What's included
The repo includes what you need to run that stack locally with Docker Compose. It does not include hosted Genwolf features such as Google AI Overviews, Google AI Mode, or live UI scraping.
Blueprint vs Hosted Genwolf
What the repo demonstrates versus what the product delivers. Hosted Genwolf is not a managed wrapper around the GitHub repo.
| Blueprint (OSS) | Hosted | |
|---|---|---|
| What it is | Reference blueprint and starting point | Managed Genwolf product |
| Purpose | Shows an approach to measuring AI visibility | Runs daily tracking and analysis for your brand |
| Engines | ChatGPT, Gemini, Perplexity via APIs | Those plus Google AI Overviews and Google AI Mode |
| Collection | Official model APIs you wire yourself | Live UI scraping, APIs where applicable |
| Operations | Your infra, keys, and maintenance | Fully managed |
| Extensibility | Fork, swap models, build integrations | Product surface with ongoing feature work |
They share a goal — understanding brand presence in AI answers. They are not the same codebase or feature set.
Deployment Model
Genwolf OSS is:
- Single-tenant by default
- Designed for internal team deployments
- Horizontally scalable on the worker layer
- Stateless on the web layer
Contributing
We welcome contributions in:
- Provider integrations
- Detection improvements
- Performance optimizations
- Documentation
Ready to explore the blueprint?
Clone the OSS repo, spin up Docker Compose, and inspect a reference visibility engine. For managed daily tracking across ChatGPT, Gemini, Perplexity, Google AI Overviews, and Google AI Mode, use hosted Genwolf.
View on GitHub