docs(attachment): state the relaxed default image limits
The default per-image limit is 50 MiB and the default per-side limit is 65,535 pixels, but the admission summaries in the attachment subsystem page and the attachment-local README still named 20 MiB and 8192 pixels. The session-controller note on the /api/file byte limit and the tool-fs note on read_image source caps also named the old 20 MiB maxImageBytes default.
This commit is contained in:
Yichen Jiang committed
1 parent
8246e568e7
commit
6dafa66dc6
12 files changed
+16
-16
No files matched your search
@@ -6,8 +6,8 @@
|
||||
en: 564dfcaaca384e23
|
||||
zh: 5d215635919b5bc4
|
||||
/durable-attachments/identity-and-verified-metadata:
|
||||
en: 3ddfd5100061dc6a
|
||||
zh: 3eedb90c8813d015
|
||||
en: 96378c208b220c6f
|
||||
zh: 7af93f6a79520732
|
||||
/durable-attachments/commit-and-verified-read-payloads:
|
||||
en: f79753272bb0ec70
|
||||
zh: 61830c41021f9274
|
||||
@@ -56,7 +56,7 @@ interface ImageAttachmentLimits {
|
||||
}
|
||||
```
|
||||
|
||||
The local backend admits at most 20 images and 200 MiB of encoded source data per message. One source may use up to 20 MiB, 64,000,000 pixels, and 8192 pixels on either side. These source limits precede the independent normalization stage, which limits the long edge to 2048 pixels and encoded data to 4 MiB by default.
|
||||
The local backend admits at most 20 images and 200 MiB of encoded source data per message. One source may use up to 50 MiB, 64,000,000 pixels, and 65,535 pixels on either side. These source limits precede the independent normalization stage, which limits the long edge to 2048 pixels and encoded data to 4 MiB by default.
|
||||
|
||||
The reference records intrinsic dimensions and encoded length so clients can lay out history without decoding first, while every authoritative read still re-checks digest, media signature, dimensions, and metadata against the object.
|
||||
|
||||
|
||||
@@ -56,7 +56,7 @@ interface ImageAttachmentLimits {
|
||||
}
|
||||
```
|
||||
|
||||
本地后端每条消息最多准入 20 张图片，源图编码数据总量不超过 200 MiB。单张源图不得超过 20 MiB、64,000,000 像素和单边 8192 像素。这些源文件限制先于独立的规范化阶段执行；该阶段默认把长边限制为 2048 像素，把编码数据限制为 4 MiB。
|
||||
本地后端每条消息最多准入 20 张图片，源图编码数据总量不超过 200 MiB。单张源图不得超过 50 MiB、64,000,000 像素和单边 65,535 像素。这些源文件限制先于独立的规范化阶段执行；该阶段默认把长边限制为 2048 像素，把编码数据限制为 4 MiB。
|
||||
|
||||
引用记录固有尺寸和编码长度，使客户端无需先解码即可排布历史记录；每次权威读取仍会根据对象重新校验摘要、媒体签名、尺寸和元数据。
|
||||
|
||||
|
||||
@@ -21,8 +21,8 @@
|
||||
en: 41bdda1ac32df558
|
||||
zh: aa5d0e325924066b
|
||||
/session-controller/session-media-references:
|
||||
en: 49a3d0e8e92017d3
|
||||
zh: 91302aa3dd6e9378
|
||||
en: 25e40cb508265c01
|
||||
zh: c8ce39a94997af00
|
||||
/session-controller/configuration:
|
||||
en: c92590fbaf75e45e
|
||||
zh: fcac67017bb0e3ab
|
||||
|
||||
@@ -86,7 +86,7 @@ References keep local Session data, scoped Contexts, and history streams alive,
|
||||
<a id="session-media-references"></a>
|
||||
## Session media references
|
||||
|
||||
`SessionMediaReferences` mounts `GET|HEAD /api/file?path=<absolute path>` on the authenticated `connection.fetch` channel when `connection`, `fs`, and `attachments` are composed. It reads ordinary files through `ctx.fs`, including temporary paths outside registered workspaces and files in remote providers. Neither directory containment nor MIME categories restrict access; `mime-types` supplies the response type, with `application/octet-stream` for unknown extensions. GET reuses `readBytes` for preflight and ongoing byte limits; HEAD reads metadata only. All files use `ctx.attachments.imageLimits.maxImageBytes` (normally 20 MiB); exceeding this limit returns 413. Responses contain the complete file, ignore Range, and carry `private, no-store`, `nosniff`, and a sandbox CSP so directly opened HTML/SVG cannot execute with the API origin. The Client rewrite lives in `ui-chat` (`AssistantMarkdown`); audio/video responses are available, while Markdown audio/video player nodes remain separate work.
|
||||
`SessionMediaReferences` mounts `GET|HEAD /api/file?path=<absolute path>` on the authenticated `connection.fetch` channel when `connection`, `fs`, and `attachments` are composed. It reads ordinary files through `ctx.fs`, including temporary paths outside registered workspaces and files in remote providers. Neither directory containment nor MIME categories restrict access; `mime-types` supplies the response type, with `application/octet-stream` for unknown extensions. GET reuses `readBytes` for preflight and ongoing byte limits; HEAD reads metadata only. All files use `ctx.attachments.imageLimits.maxImageBytes` (normally 50 MiB); exceeding this limit returns 413. Responses contain the complete file, ignore Range, and carry `private, no-store`, `nosniff`, and a sandbox CSP so directly opened HTML/SVG cannot execute with the API origin. The Client rewrite lives in `ui-chat` (`AssistantMarkdown`); audio/video responses are available, while Markdown audio/video player nodes remain separate work.
|
||||
|
||||
GUI model selection requires the exact provider/model pair in the available catalog; unavailable choices reject with `session/model-unavailable`. Prompt admission retains the saved route without a catalog availability gate, so request execution reports missing credentials or unavailable models. `initializeDefaultModel()` saves the first available account model after login when no other provider has a configured API key; credential checks use configured references independently of model availability. A provider with no available models rejects initialization with `session/provider-models-unavailable`. Availability never substitutes another model or rewrites a Session selection.
|
||||
|
||||
|
||||
@@ -86,7 +86,7 @@ Client 的首次 `follow`、重连首屏与 `loadOlder()` 至少请求 50 条以
|
||||
<a id="session-media-references"></a>
|
||||
## 会话媒体引用
|
||||
|
||||
当 `connection`、`fs` 与 `attachments` 均被组合时，`SessionMediaReferences` 在鉴权 `connection.fetch` 通道上挂载 `GET|HEAD /api/file?path=<绝对路径>`。它通过 `ctx.fs` 读取普通文件，包括已注册工作区之外的临时路径与远程提供方中的文件。目录包含关系与 MIME 类别均不限制访问；`mime-types` 提供响应类型，未知扩展名使用 `application/octet-stream`。GET 复用 `readBytes` 执行读取前及读取中的字节限制；HEAD 只读取元数据。所有文件均使用 `ctx.attachments.imageLimits.maxImageBytes`（通常为 20 MiB）；超过此上限返回 413。响应包含完整文件，忽略 Range，并携带 `private, no-store`、`nosniff` 与沙箱 CSP，使直接打开的 HTML/SVG 无法以 API 源身份执行脚本。客户端重写位于 `ui-chat`（`AssistantMarkdown`）；音视频文件响应已可用，Markdown 音视频播放器节点仍是独立工作。
|
||||
当 `connection`、`fs` 与 `attachments` 均被组合时，`SessionMediaReferences` 在鉴权 `connection.fetch` 通道上挂载 `GET|HEAD /api/file?path=<绝对路径>`。它通过 `ctx.fs` 读取普通文件，包括已注册工作区之外的临时路径与远程提供方中的文件。目录包含关系与 MIME 类别均不限制访问；`mime-types` 提供响应类型，未知扩展名使用 `application/octet-stream`。GET 复用 `readBytes` 执行读取前及读取中的字节限制；HEAD 只读取元数据。所有文件均使用 `ctx.attachments.imageLimits.maxImageBytes`（通常为 50 MiB）；超过此上限返回 413。响应包含完整文件，忽略 Range，并携带 `private, no-store`、`nosniff` 与沙箱 CSP，使直接打开的 HTML/SVG 无法以 API 源身份执行脚本。客户端重写位于 `ui-chat`（`AssistantMarkdown`）；音视频文件响应已可用，Markdown 音视频播放器节点仍是独立工作。
|
||||
|
||||
GUI 模型选择要求确切提供方／模型对出现在可用目录中；不可用的选择以 `session/model-unavailable` 拒绝。提示词准入保留已保存的路由，不按目录可用性阻断发送，由请求执行报告凭据缺失或模型不可用。`initializeDefaultModel()` 在账号登录后、其他提供方均未配置 API key 时，将第一个可用账号模型保存为默认模型；凭据检查使用已配置的引用，不依赖模型是否可用。提供方没有可用模型时，初始化以 `session/provider-models-unavailable` 拒绝。可用性变化不会替换模型或改写会话选择。
|
||||
|
||||
|
||||
@@ -36,8 +36,8 @@
|
||||
en: dcd9857653645420
|
||||
zh: a27259a09f17465d
|
||||
/deepseek-ai-dsh-attachment-local/understand-the-implementation/write-and-read-paths:
|
||||
en: 609ccc784a87b770
|
||||
zh: bc702582cf2b3316
|
||||
en: 62ea104a00a7fdb8
|
||||
zh: b08575274ca244b4
|
||||
/deepseek-ai-dsh-attachment-local/understand-the-implementation/source-map:
|
||||
en: 05af85aa85934d01
|
||||
zh: 1c89e80d496dc0a3
|
||||
|
||||
@@ -83,7 +83,7 @@ This section explains the durability and verification design behind the storage,
|
||||
|
||||
Objects land at `<DSH_HOME>/attachments/v1/objects/<sha256-prefix>/<sha256>`; equal bytes deduplicate to one object and one `sha256:` id. Before the first write, the process syncs every ancestor directory of the home down to the filesystem root once, so a directory another process created but has not yet synced is never mistaken for a safe boundary. Writes then stage bytes in `v1/tmp`, sync the temporary file, publish with an atomic exclusive hard link, and sync the publication directories — on Windows, filesystem metadata journaling owns entry durability. Once the save resolves, the reported reference is durable.
|
||||
|
||||
Admission accepts up to 20 images and 200 MiB of source bytes per message; one source may use up to 20 MiB, 64 million pixels, and 8192 pixels per side. It applies orientation, removes metadata and color profiles, and normalizes under a 2048×2048 total-pixel budget, an 8192-pixel long edge, and a 4 MiB encoded-byte target. Extreme aspect ratios therefore retain their short-edge resolution. Clean single-frame 8-bit sRGB/sRGBA PNG, JPEG, or WebP input already within those limits passes through byte-identically; GIF, animation, metadata, orientation, 16-bit PNG, and incompatible color spaces force conversion.
|
||||
Admission accepts up to 20 images and 200 MiB of source bytes per message; one source may use up to 50 MiB, 64 million pixels, and 65,535 pixels per side. It applies orientation, removes metadata and color profiles, and normalizes under a 2048×2048 total-pixel budget, an 8192-pixel long edge, and a 4 MiB encoded-byte target. Extreme aspect ratios therefore retain their short-edge resolution. Clean single-frame 8-bit sRGB/sRGBA PNG, JPEG, or WebP input already within those limits passes through byte-identically; GIF, animation, metadata, orientation, 16-bit PNG, and incompatible color spaces force conversion.
|
||||
|
||||
Request versions live below `<DSH_HOME>/cache/attachments/request-images/`, resolved by `dshCachePath`; an explicit `dshHome` setting applies to both cache and durable storage. Clearing this cache between requests preserves durable attachments, and later reads regenerate the variants. `readImageRequest` scales without enlargement to the route-chosen target, resizing by the long edge only so the encoder derives the short edge as the route predicts, then applies a separate encoded-byte target through the same alpha routing and quality ladder. Its cache identity includes the attachment id, transform version, target dimensions, byte target, and fixed encoder settings; cached bytes are header-probed for format, 8-bit sRGB/sRGBA, dimensions, and alpha facts, and a mismatch regenerates the entry. Concurrent callers share one transform and cache write, while cancellation stops shared work only when no waiter remains. `imageHostPath` derives the normalized object's host path, and the mounted filesystem may map that path into its execution world without writing it to durable history.
|
||||
|
||||
|
||||
@@ -83,7 +83,7 @@ kind: "package-reference"
|
||||
|
||||
对象存放在 `<DSH_HOME>/attachments/v1/objects/<sha256-prefix>/<sha256>`；相同字节会去重为同一个对象和同一个 `sha256:` 标识符。首次写入前，进程会把 home 的每个祖先目录逐级同步到文件系统根目录，因此绝不会把另一个进程已创建但尚未同步的目录误认为安全边界。随后，写入过程把字节暂存到 `v1/tmp`、同步临时文件、以原子且排他的硬链接发布，并同步发布目录——在 Windows 上，文件系统元数据日志负责目录项持久性。保存操作完成后，返回的引用已具备持久性。
|
||||
|
||||
准入允许每条消息最多 20 张图片与 200 MiB 源字节；单个源图最多 20 MiB、6400 万像素与单边 8192 像素。系统应用方向、移除元数据与色彩配置，并把规范化结果限制在 2048×2048 总像素预算、8192 像素长边和 4 MiB 编码字节目标内，因此，即使宽高比极端，图片也会保留短边分辨率。已经满足限制的干净、单帧、8-bit sRGB/sRGBA PNG、JPEG 或 WebP 会逐字节直通；GIF、动画、元数据、方向、16-bit PNG 与不兼容色彩空间会触发转换。
|
||||
准入允许每条消息最多 20 张图片与 200 MiB 源字节；单个源图最多 50 MiB、6400 万像素与单边 65,535 像素。系统应用方向、移除元数据与色彩配置，并把规范化结果限制在 2048×2048 总像素预算、8192 像素长边和 4 MiB 编码字节目标内，因此，即使宽高比极端，图片也会保留短边分辨率。已经满足限制的干净、单帧、8-bit sRGB/sRGBA PNG、JPEG 或 WebP 会逐字节直通；GIF、动画、元数据、方向、16-bit PNG 与不兼容色彩空间会触发转换。
|
||||
|
||||
请求版本位于由 `dshCachePath` 解析的 `<DSH_HOME>/cache/attachments/request-images/`；显式 `dshHome` 设置同时适用于缓存与持久存储。在两次请求之间清空此缓存会保留持久附件，后续读取会重新生成请求版本。`readImageRequest` 在不放大的前提下缩放到路由选定的目标尺寸，缩放只按长边给定，短边由编码器按路由预测的方式推出，随后通过相同的 alpha 路由与质量阶梯应用独立编码字节目标。缓存身份包含附件 id、变换版本、目标尺寸、字节目标与固定编码参数；缓存字节会先通过文件头探测格式、8-bit sRGB/sRGBA、尺寸与 alpha 信息，不匹配时重新生成。并发调用方共享一次变换与缓存写入，且只在没有等待方时由取消停止共享工作。`imageHostPath` 派生规范化对象的宿主路径，挂载的文件系统可以把该路径映射进执行世界，而不会写入持久历史。
|
||||
|
||||
|
||||
@@ -135,8 +135,8 @@
|
||||
en: f5d5f97a153f7c70
|
||||
zh: 3d6bc05b9eadfbc2
|
||||
/deepseek-ai-dsh-tool-fs/known-limitations-and-deferred-work:
|
||||
en: ce7a91d41d2f7597
|
||||
zh: d4a3c905e124fd7f
|
||||
en: 28aadb772cdbc497
|
||||
zh: a50e3bb6b4a8f645
|
||||
/deepseek-ai-dsh-tool-fs/known-limitations-and-deferred-work/dev-note:
|
||||
en: e67f633dc6fe7773
|
||||
zh: 3bb0c20c93f16f98
|
||||
@@ -250,7 +250,7 @@ These limits define when the tool suite is a poor fit or needs special operation
|
||||
- **No model-facing directory listing ships** — `ctx.fs.listDir` serves provider code such as skill discovery, while the sibling `dsh-tool-fs-search` package supplies ripgrep-backed `glob` and `grep` rather than extending the filesystem seam.
|
||||
- **`read` handles UTF-8 text files only** — images use the separate `read_image` tool; PDF, audio, and video remain deferred. A directory target is `FS_NOT_REGULAR_FILE`.
|
||||
- **Extension-declared media type** — an extension selects the declared type and the attachment store's magic-byte validation stays authoritative; a correctly formatted image under a wrong extension is refused with the rename remedy rather than sniffed. Only a path with no extension is identified from its file signature.
|
||||
- **Object paths re-enter source admission** — `read_image` on a normalized attachment object re-admits its bytes as a new source, so a deployment whose `maxImageBytes`/`maxMessageImageBytes` sit below the normalized-image byte budget can refuse an object path that `ctx.attachments.readImage` still serves; shipped defaults keep the normalized budget (4 MiB) far under the source caps (20 MiB).
|
||||
- **Object paths re-enter source admission** — `read_image` on a normalized attachment object re-admits its bytes as a new source, so a deployment whose `maxImageBytes`/`maxMessageImageBytes` sit below the normalized-image byte budget can refuse an object path that `ctx.attachments.readImage` still serves; shipped defaults keep the normalized budget (4 MiB) far under the source caps (50 MiB).
|
||||
- **Inline image preview rides the UI composition** — the tool-result card renders the image through the browser's `tool.call.images` slot, which the attachment presentation plugin fills; a UI without that plugin shows the result's envelope text instead.
|
||||
- **No attachment-region tool** — an agent may crop an image through another available tool when it has a filesystem path; a pasted or dragged image without a path cannot be re-read at higher resolution.
|
||||
- **No timeout surface** — `read`/`write`/`edit` take no timeout argument and declare no timeout budget; cancellation rides `exec.signal` only ([provider rationale](../README.md)).
|
||||
|
||||
@@ -250,7 +250,7 @@ Read a file before editing it (the default fs-observation-policy requires it), u
|
||||
- **未交付面向模型的目录列表工具**：`ctx.fs.listDir` 服务于 skill（技能）发现等提供方代码，同级 `dsh-tool-fs-search` 包则提供基于 ripgrep 的 `glob` 与 `grep`，而不是扩展文件系统 seam。
|
||||
- **`read` 只处理 UTF-8 文本文件**：图像使用独立的 `read_image` 工具；PDF、音频和视频仍延期处理。目录目标为 `FS_NOT_REGULAR_FILE`。
|
||||
- **媒体类型按扩展名声明**：扩展名选择声明类型，附件存储的魔数校验保持权威；扩展名错误但格式正确的图像会得到改名修复提示，而不是被嗅探接受。只有没有扩展名的路径按文件签名识别格式。
|
||||
- **对象路径重新走源准入**：对规范化附件对象调用 `read_image` 会把其字节作为新来源重新准入，因此把 `maxImageBytes`/`maxMessageImageBytes` 配置得低于规范化图片字节预算的部署可能拒绝 `ctx.attachments.readImage` 仍可读取的对象路径；默认配置下规范化预算（4 MiB）远低于源上限（20 MiB）。
|
||||
- **对象路径重新走源准入**：对规范化附件对象调用 `read_image` 会把其字节作为新来源重新准入，因此把 `maxImageBytes`/`maxMessageImageBytes` 配置得低于规范化图片字节预算的部署可能拒绝 `ctx.attachments.readImage` 仍可读取的对象路径；默认配置下规范化预算（4 MiB）远低于源上限（50 MiB）。
|
||||
- **内嵌图像预览依赖 UI 组合**：工具结果卡片经由浏览器的 `tool.call.images` 槽位渲染图像，由附件呈现插件填充；未组合该插件的 UI 改为显示结果的信封文本。
|
||||
- **没有附件区域工具**：agent 在拥有文件系统路径时可以通过其他可用工具裁剪图片；没有路径的粘贴或拖入图片无法按更高分辨率重新读取。
|
||||
- **没有超时接口**：`read`/`write`/`edit` 不接受超时参数，也不声明超时预算；取消只通过 `exec.signal` 传递（见[提供方理由](../README.zh.md)）。
|
||||
|
||||
Reference in new issue
Block a user