Skip to content
Kumiko

ai-conversation

Enterprise Part of the commercial Kumiko Enterprise offering — not in the open framework.

Persisted multi-turn refinement dialog for the Designer chat. The ai-conversation:write:send write-handler loads a conversation’s prior turns, routes the prompt to the tenant’s configured LLM provider (suggestPatch on the first turn, refinePattern with history on follow-ups), and appends the user and assistant turns as immutable designer-conversation-turn events — provider errors are returned to the caller but never persisted. ai-conversation:query:history returns a conversation’s turns for rendering. The system prompt, model, effort and token budget of each route live in the prompt-store under ai-conversation:suggest-patch (first turn) and ai-conversation:refine-pattern (follow-ups), with the shipped texts as defaults. Requires ai-foundation for the LLM provider wiring and prompt-store for the prompts.

Build it yourself

Section titled “Build it yourself”

This Enterprise implementation is optional. If you need a narrower or domain-specific version, you can build an app-owned feature, handler, or job with Kumiko’s public extension points.

Use this package when you want the maintained implementation, its existing contracts, and the surrounding operational behavior. A custom implementation is not drop-in compatible automatically: your app owns its authorization, validation, integrations, tests, security review, and production operations.

Start with Features and composition and Custom write handlers.

How it fits

Section titled “How it fits”

What this feature needs to run (Requires, top) and the write commands this feature provides (Provides, bottom).

flowchart TB
  n_ai_conversation["ai-conversation"]
  subgraph how_reqs["Requires"]
    n_ai_foundation["ai-foundation"]
    n_prompt_store["prompt-store"]
    n_tenant["tenant"]
  end
  subgraph how_provides["Provides"]
    n_cmd_ai_conversation_write_send(["send"])
  end
  n_ai_foundation --> n_ai_conversation
  n_prompt_store --> n_ai_conversation
  n_tenant --> n_ai_conversation
  n_ai_conversation --> n_cmd_ai_conversation_write_send

Provides — write commands this feature registers (dispatch them through the command bus):

  • ai-conversation:write:send

Dependencies

Section titled “Dependencies”

Extensions & cross-feature APIs

Section titled “Extensions & cross-feature APIs”
  • Registers extension: promptTemplate → ai-conversation:suggest-patch, promptTemplate → ai-conversation:refine-pattern