Package com.axelor.ai.chat.plugin

package com.axelor.ai.chat.plugin
  • Class
    Description
    AgentResolverPlugin
    Resolves the agent/model/provider config from the DB and builds a real LangChain4jChatProvider so the LLM actually responds.
    ContextPlacement
    Places per-turn context beside the message it serves, not at the top of the prompt.
    ContextWindowPlugin
    Manages the LLM context window by transparently trimming message history when it exceeds the model's input token limit.
    ConversationSummarizer
    Condenses the stretch of conversation that is rolling out of the context window.
    FieldContextPlugin
    Names the fields a user message was sent with.
    InputGuardPlugin
    Defends against prompt injection by injecting security rules into the conversation.
    MentionContextPlugin
    Resolves structured user mentions into trusted model context.
    PromptedConversationSummarizer
    Summarizes by asking a model, on whatever provider the conversation already uses.
    RiskScreenPlugin
    The third layer of the confused-deputy defence: before the target agent consumes a discussion containing other members' text, an internal risk-assessor agent screens the fresh span (the human messages since the agent's last reply) for manipulation attempts.
    SessionNotesPlugin
    Incrementally builds structured session notes during a conversation.
    SkillPlugin
    Keeps the agent's skill catalog in the system prompt, and backfills skill content the history does not carry.
    SubjectPlugin
    Pins the record a conversation is about.
    SubjectPlugin.PinnedFields
    Pinned fields resolved once per model and path for one reader: a chat listing shows many chats pinned on the same few fields, and each resolve reads permissions and metadata.
    TitleGeneratorPlugin
    Titles a new chat once its first answer is in: a short model call over the user's message and the reply, run after the stream ends so the reader never waits on it, and announced to the chat's room like any other change.
    TokenEstimator
    Estimates token counts for UIMessage lists using provider-specific tokenizers when available, falling back to a character-based heuristic for providers without public tokenizers.