Hi @trypromptly maintainers,
I’m Vivek from Memcode. In a visual LLM workflow builder, users may want a later workflow to access a few durable preferences or decisions from an earlier one. That is different from persisting every intermediate node output or making a knowledge base from all workflow data.
Would a Memcode MCP/tool connector be useful in LLMStack? A user would explicitly configure and invoke it, authorize through OAuth/DCR and PKCE, and decide what to save. The integration would keep user data isolated and never send prompts, secrets, or workflow payloads automatically.
If this aligns with the product, which node or integration surface should a first PR use?
Hi @trypromptly maintainers,
I’m Vivek from Memcode. In a visual LLM workflow builder, users may want a later workflow to access a few durable preferences or decisions from an earlier one. That is different from persisting every intermediate node output or making a knowledge base from all workflow data.
Would a Memcode MCP/tool connector be useful in LLMStack? A user would explicitly configure and invoke it, authorize through OAuth/DCR and PKCE, and decide what to save. The integration would keep user data isolated and never send prompts, secrets, or workflow payloads automatically.
If this aligns with the product, which node or integration surface should a first PR use?