From 78321e6a8a0b122d475db65db510d68fa13b1795 Mon Sep 17 00:00:00 2001 From: wang-qisen Date: Thu, 10 Sep 2026 18:05:46 +0800 Subject: [PATCH] docs(auto-memory): clarify history rendering scope for direct inputs --- docs/en/auto_memory.md | 6 +++--- docs/zh/auto_memory.md | 2 +- 2 files changed, 4 insertions(+), 4 deletions(-) diff --git a/docs/en/auto_memory.md b/docs/en/auto_memory.md index b30f497f..76a1064c 100644 --- a/docs/en/auto_memory.md +++ b/docs/en/auto_memory.md @@ -98,9 +98,9 @@ orientation correction and format conversion when needed). Original caller messa ### Direct multimodal input (default when enabled) -Auto Memory builds an AgentScope `UserMsg` with text and image `DataBlock` values interleaved within the rendered -conversation history. It keeps their original order and each message's speaker and timestamp boundaries, including -image-only messages and repeated block IDs. This describes the input ReMe passes to the wrapper; the selected +With its standard history rendering, Auto Memory builds an AgentScope `UserMsg` with text and image `DataBlock` values +interleaved within the conversation. It keeps their original order and each message's speaker and timestamp boundaries, +including image-only messages and repeated block IDs. This describes the input ReMe passes to the wrapper; the selected AgentScope formatter determines how that input is represented in provider requests. The memory Agent extracts facts from text and images together, without separate caption calls. This is one Agent workflow, not necessarily one API request: tool use can trigger further model turns that still include the images. diff --git a/docs/zh/auto_memory.md b/docs/zh/auto_memory.md index 9e35abf0..53ca2524 100644 --- a/docs/zh/auto_memory.md +++ b/docs/zh/auto_memory.md @@ -90,7 +90,7 @@ reme auto_memory session_id=session-a include_images=true messages='[...]' ### 直接多模态输入（开启图像后的默认模式） -Auto Memory 构建 AgentScope `UserMsg`，在渲染后的会话历史中按原顺序交错放置文本与图像 `DataBlock`， +使用默认的会话渲染时，Auto Memory 构建 AgentScope `UserMsg`，在会话历史中按原顺序交错放置文本与图像 `DataBlock`， 保留每条消息的说话人和时间边界，也支持只有图像的消息及重复 block ID。 这里描述的是 ReMe 传给 wrapper 的输入；实际 provider 请求中的表示方式由所选 AgentScope formatter 决定。 记忆 Agent 同时读取文本与图像，不单独调用 caption 模型。这是一次 Agent 工作流，不保证只有一次 API 请求：