feat: 手动触发 DMN、备忘录回读与记忆列表分页

[AmriaLin]

变更内容

1. /memory dmn — 手动触发潜意识调度

• SubconsciousRunner.trigger_now()：取消已排的 date 任务，取消上一次手动任务，用 create_task 立即跑一轮
• 任务引用保存在 _manual_task，避免被 GC 回收；stop() 一并取消
• 命令入口 /memory dmn（别名 run、触发），仅超级用户可用，runner 未启动时返回明确提示

2. /memory memo — 查看备忘录

• 以合并转发形式发送，直接复用 amrita.utils.send.send_forward_msg
• 内容按行装箱，单行超过 600 字才按句切开，不从句子中间截断
• 接口不可用时退回纯文本分段发送，不影响命令完成

3. read_memo / append_memo 工具

备忘录常驻注入系统提示，但 LLM 拿不到原文，也无法只补一条规则。

• read_memo：只读，返回全文、字数与 empty 标志
• append_memo：在末尾追加一段，与原文以换行分隔，仍受 memo_max_chars 上限约束；整篇重写仍走 update_memo

4. 记忆列表分页

• AsyncUserMemory.get_all_notes 增加 limit / offset，透传给 chroma 的 get
• list_memory 工具新增 offset，limit 默认 10、上限 50，返回 total / has_more / next_offset，不再整库拉取后切片
• /memory user|group list [页码] 每页 10 条，显示页码并提示下一页

说明

• 文档只更新了 README（仓库内没有 docs 与 docs/zh）
• 命令与工具注册在启动时完成，需要重启 bot 生效

Summary by Sourcery

Enable on-demand subconscious runs, memo access and incremental updates, and efficient paginated memory browsing.

New Features:

• Add privileged commands for manually triggering subconscious scheduling and viewing the current-scope memo.
• Add memo read and append tools for inspecting and incrementally updating memo content.
• Add paginated memory listing for chat commands and LLM tools.

Enhancements:

• Track manually triggered subconscious runs and include their token usage in target-user statistics.
• Provide fallback text delivery when memo forwarding is unavailable.

Build:

• Bump the plugin version to 0.7.0.

Documentation:

• Update the README with the new memo tools and memory commands.

[sourcery-ai]

Sorry @AmriaLin, you've used your own review budget of 250,000 diff characters for the last 7 days.

You can request another review in 24 minutes by commenting @sourcery-ai review. Upgrade to get a review now.

[sourcery-ai]

Reviewer's Guide

本 PR 增加了潜意识调度的超级用户手动触发、备忘录的合并转发与全文回读工具，并将记忆列表从本地切片改为基于 Chroma limit/offset 的分页，同时更新命令和工具文档。

Sequence diagram for manual subconscious scheduling

sequenceDiagram
    actor Superuser
    participant Matcher
    participant Runner as SubconsciousRunner
    participant Scheduler
    participant Task as asyncio.Task

    Superuser->>Matcher: /memory dmn
    Matcher->>Runner: trigger_now()
    Runner->>Scheduler: remove_job(_job_id)
    Runner->>Task: asyncio.create_task(_run())
    Runner-->>Matcher: trigger_now() completes
    Matcher-->>Superuser: 已触发，潜意识将在后台跑一轮
    Task->>Runner: _run()

Loading

Sequence diagram for memo viewing and fallback delivery

sequenceDiagram
    actor User
    participant Matcher
    participant Memo as get_memo()
    participant Bot

    User->>Matcher: /memory memo
    Matcher->>Memo: get_memo(scope_id)
    Memo-->>Matcher: memo content
    Matcher->>Bot: send_group_forward_msg or send_private_forward_msg
    alt forward API fails
        Bot-->>Matcher: exception
        Matcher->>Matcher: send(content)
    end
    Matcher-->>User: memo chunks labeled by segment

Loading

Sequence diagram for paginated memory listing

sequenceDiagram
    actor User
    participant Matcher
    participant Memory as AsyncUserMemory
    participant Chroma

    User->>Matcher: /memory user list page
    Matcher->>Memory: count_user_notes(partition_id)
    Memory->>Chroma: count()
    Chroma-->>Memory: total
    Matcher->>Memory: get_all_notes(limit=10, offset=page offset)
    Memory->>Chroma: get(limit, offset)
    Chroma-->>Memory: page of notes
    Memory-->>Matcher: IDs, documents, metadatas
    Matcher-->>User: page results and next-page hint

Loading

File-Level Changes

Change	Details	Files
增加潜意识调度的一次性手动触发能力，并通过命令入口限制超级用户访问。	取消已排程任务和正在执行的手动任务后创建后台任务 保存手动任务引用，并在停止 runner 时取消 新增 /memory dmn、run、触发 命令及未启动提示	amrita_plugin_memory/rethinking/runner.py amrita_plugin_memory/matchers.py
新增备忘录命令展示与 LLM 原文读取能力，并为转发失败提供降级路径。	按 600 字拆分备忘录并按群聊或私聊调用合并转发 API 转发失败时回退为纯文本发送 新增 read_memo 工具返回全文、字数和空内容标志	amrita_plugin_memory/matchers.py amrita_plugin_memory/tools.py README.md
将记忆查询改为后端原生分页，并同步扩展工具和用户命令的分页元数据。	为 Chroma 查询透传 limit 和 offset list_memory 增加参数校验、总数及下一页信息，避免整库加载 `/memory user	group list` 改为每页 10 条并显示页码和下一页提示

---

Tips and commands

Interacting with Sourcery

• Trigger a new review: Comment @sourcery-ai review on the pull request.
• Continue discussions: Reply directly to Sourcery's review comments.
• Generate a GitHub issue from a review comment: Ask Sourcery to create an
  issue from a review comment by replying to it. You can also reply to a
  review comment with @sourcery-ai issue to create an issue from it.
• Generate a pull request title: Write @sourcery-ai anywhere in the pull
  request title to generate a title at any time. You can also comment
  @sourcery-ai title on the pull request to (re-)generate the title at any time.
• Generate a pull request summary: Write @sourcery-ai summary anywhere in
  the pull request body to generate a PR summary at any time exactly where you
  want it. You can also comment @sourcery-ai summary on the pull request to
  (re-)generate the summary at any time.
• Generate reviewer's guide: Comment @sourcery-ai guide on the pull
  request to (re-)generate the reviewer's guide at any time.
• Resolve all Sourcery comments: Comment @sourcery-ai resolve on the
  pull request to resolve all Sourcery comments. Useful if you've already
  addressed all the comments and don't want to see them anymore.
• Dismiss all Sourcery reviews: Comment @sourcery-ai dismiss on the pull
  request to dismiss all existing Sourcery reviews. Especially useful if you
  want to start fresh with a new review - don't forget to comment
  @sourcery-ai review to trigger a new review!

Customizing Your Experience

Access your dashboard to:

• Enable or disable review features such as the Sourcery-generated pull request
  summary, the reviewer's guide, and others.
• Change the review language.
• Add, remove or edit custom review instructions.
• Adjust other review settings.

Getting Help

• Contact our support team for questions or feedback.
• Visit our documentation for detailed guides and information.
• Keep in touch with the Sourcery team by following us on X/Twitter, LinkedIn or GitHub.