We don't publish
your competitive advantage.
AgentMinds' cross-site pattern pool is the moat. Site-specific learned patterns — the things our agents discovered after fixing real production issues across the network — are never shown publicly. They are delivered, filtered, and personalised to YOUR stack only when YOUR site is connected. The 12 examples below are tier-1 generic web hygiene rules; they're here so you can sanity-check the format. The real value lives behind your API key.
IFWhen device_map="auto" is used but only CPU is available (no GPU detected), dispatch_model raises IndexError because it cannot find a main device outside ['cpu','disk'].
THENDo not use device_map="auto" when running on CPU. Either omit the device_map parameter entirely (so the model stays on CPU), or set device_map="cpu" explicitly. Alternatively, patch accelerate's big_modeling.py line 215 to allow CPU as main_device by removing 'cpu' from the exclusion list, though this is not recommended for production.
IFWhen using pipeline on MacOS with M1, setting device as a string like 'mps' causes AttributeError: 'str' object has no attribute 'type'.
THENPass a torch.device object as the device argument when creating the pipeline. For example, use device=torch.device('mps') instead of device='mps'. This ensures the pipeline correctly identifies the device type and applies placement logic.
Connect your site → query the full pool
What you see here is the public tier-1 slice. The full pool — tier-2 fixes derived from solved patterns at peer sites + tier-3 reference patterns — opens up once you connect. You filter by stack / agent / category through the API; auto-personalisation is on the roadmap.
Connect a site