Bitsec is an Ecosystem for AI-powered code vulnerability detection. It finds and fixes code vulnerabilities for subnet codebases and smart contracts. As the project grows, Bitsec will expand coverage to other types of codebases and vulnerabilities.
| Docs | Resources |
|
Mining and Validator Guide Contributor Guide |
SN60-Powered Applications Bittensor Introduction Bittensor Guides BitsecAI twitter / X |
Since testnet, we have been evaluating subnet incentive mechanisms with critical, high, medium findings in mainnet.
These will be run regularly and summaries posted on X, with detailed findings on our website.
The subnet incentivizes miners to develop models and workflows to find and fix code vulnerabilities. Applications allow paying users access to Bitsec's network of miners.
Here are two sample applications on the roadmap:
- Bitsec Scanner: Give it a github repo, the request is sent to our community of miners to find exploits.
- Bitsec Hunter: Participate in active bug bounty programs.
- Miners: Run machines and workflows that find software exploits. The system is amenable to a wide variety of exploit detection techniques, including machine learning, fine tuned models, AI agent frameworks, static analysis, etc.
- Research Integration: We update our detection models and methodologies in response to emerging academic research, offering resources like training data sets, training code, model weights to our community.
- Validators: Responsible for challenging miners with a balanced mix of synthetic and real codebases, drawn from a diverse pool of sources.
- Resource Expansion: We are committed to enhancing the validators' capabilities by increasing the diversity and volume of the codebase population, which supports rigorous testing and validation processes.
Ask and participate in the Bittensor discord, look for the channel for subnet 60.
