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 using Langfuse's OpenAI wrapper with `store=True` for model distillation, the `metadata` parameter intended for OpenAI is intercepted by Langfuse and never forwarded to OpenAI, causing stored completions to lack filtering metadata.
THENUpgrade to Langfuse Python SDK v2.56.0 or later, which passes `metadata` to OpenAI when `store=True`. If upgrading is not possible, use the native OpenAI client directly for operations requiring `metadata` and `store`, or introduce a custom parameter (e.g., `openai_metadata`) and manually map it in the call.
IFWhen using Langfuse Python wrapper for OpenAI, the `metadata` parameter intended for OpenAI chat completions (e.g., for model distillation with `store=True`) is intercepted by Langfuse for its own usage, preventing the metadata from reaching OpenAI.
THENModify the Langfuse wrapper to differentiate between Langfuse metadata and OpenAI metadata, e.g., by introducing an `openai_metadata` parameter that is forwarded to the OpenAI API internally. Alternatively, make direct OpenAI API calls for operations that require the `metadata` field, bypassing the Langfuse wrapper. A fix has been proposed in a linked PR to pass metadata to OpenAI when model distillation is used.
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