@oadank/dsh-input-tools 0.3.23 → 0.3.24

package/lib/index.js

@@ -310,6 +310,105 @@ async function saveVoiceFile(root, data, mediaType, durationMs) {
   }
 }
 
+// ──────────────────────────────────────────────────────────────
+// 图片对象存储（内容寻址，与语音/用户附件同池：DSH_HOME/attachments/v1/objects）
+// [2026-08-23] send_image 工具：agent 主动发图，落盘为独立 image/reply 事件。
+// ──────────────────────────────────────────────────────────────
+const MAX_IMAGE_BYTES = 30 * 1024 * 1024
+
+/** 从文件路径嗅探图片媒体类型（send_image 用）。 */
+function sniffImageType(path) {
+  const ext = (path.split('.').pop() ?? '').toLowerCase()
+  if (ext === 'png') return 'image/png'
+  if (ext === 'jpg' || ext === 'jpeg') return 'image/jpeg'
+  if (ext === 'gif') return 'image/gif'
+  if (ext === 'webp') return 'image/webp'
+  return undefined
+}
+
+/** 最佳努力解析图片内禀尺寸（png/jpeg/gif/webp），失败返回 1×1（RPC schema 要求正数）。 */
+function readImageSize(data) {
+  try {
+    if (data.length >= 24 && data[0] === 0x89 && data[1] === 0x50 && data[2] === 0x4e && data[3] === 0x47) {
+      const w = (data[16] << 24) | (data[17] << 16) | (data[18] << 8) | data[19]
+      const h = (data[20] << 24) | (data[21] << 16) | (data[22] << 8) | data[23]
+      return { width: Math.max(1, w), height: Math.max(1, h) }
+    }
+    if (data.length >= 10 && data[0] === 0x47 && data[1] === 0x49 && data[2] === 0x46) {
+      const w = data[6] | (data[7] << 8)
+      const h = data[8] | (data[9] << 8)
+      return { width: Math.max(1, w), height: Math.max(1, h) }
+    }
+    if (data.length >= 4 && data[0] === 0xff && data[1] === 0xd8) {
+      let i = 2
+      while (i + 9 < data.length) {
+        if (data[i] !== 0xff) { i += 1; continue }
+        const marker = data[i + 1]
+        if (marker === 0xd9 || marker === 0xda) break
+        if (marker >= 0xc0 && marker <= 0xcf && marker !== 0xc4 && marker !== 0xc8 && marker !== 0xcc) {
+          const h = (data[i + 5] << 8) | data[i + 6]
+          const w = (data[i + 7] << 8) | data[i + 8]
+          return { width: Math.max(1, w), height: Math.max(1, h) }
+        }
+        const len = (data[i + 2] << 8) | data[i + 3]
+        i += 2 + len
+      }
+    }
+    if (
+      data.length >= 16
+      && data[0] === 0x52 && data[1] === 0x49 && data[2] === 0x46 && data[3] === 0x46
+      && data[8] === 0x57 && data[9] === 0x45 && data[10] === 0x42 && data[11] === 0x50
+    ) {
+      const fourcc = String.fromCharCode(data[12], data[13], data[14], data[15])
+      if (fourcc === 'VP8X' && data.length >= 30) {
+        const w = (data[24] | (data[25] << 8) | (data[26] << 16)) + 1
+        const h = (data[27] | (data[28] << 8) | (data[29] << 16)) + 1
+        return { width: w, height: h }
+      }
+      if (fourcc === 'VP8L' && data.length >= 25) {
+        const b = data[21] | (data[22] << 8) | (data[23] << 16) | (data[24] << 24)
+        return { width: (b & 0x3fff) + 1, height: ((b >> 14) & 0x3fff) + 1 }
+      }
+      if (fourcc === 'VP8 ' && data.length >= 32) {
+        const w = data[26] | (data[27] << 8) | ((data[28] & 0x3f) << 16)
+        const h = data[29] | (data[30] << 8) | ((data[31] & 0x3f) << 16)
+        return { width: Math.max(1, w), height: Math.max(1, h) }
+      }
+    }
+  } catch { /* 忽略 */ }
+  return { width: 1, height: 1 }
+}
+
+async function saveImageFile(root, data, mediaType) {
+  if (data.byteLength > MAX_IMAGE_BYTES) {
+    throw new Error(`Image object exceeds the ${MAX_IMAGE_BYTES}-byte limit.`)
+  }
+  const sha256 = createHash('sha256').update(data).digest('hex')
+  const bucket = join(root, 'objects', sha256.slice(0, 2))
+  const target = objectPath(root, sha256)
+  await mkdir(bucket, { recursive: true, mode: 0o700 })
+  let handle
+  try {
+    handle = await open(target, constants.O_CREAT | constants.O_EXCL | constants.O_WRONLY, 0o600)
+    await handle.writeFile(data)
+    await handle.close()
+    handle = undefined
+  } catch (error) {
+    if (handle !== undefined) await handle.close().catch(() => {})
+    if (!(error instanceof Error && 'code' in error && error.code === 'EEXIST')) {
+      throw new Error(`Unable to persist image object: ${String(error)}`, { cause: error })
+    }
+  }
+  const { width, height } = readImageSize(data)
+  return {
+    attachmentId: `sha256:${sha256}`,
+    mediaType,
+    bytes: data.byteLength,
+    width,
+    height,
+  }
+}
+
 // ──────────────────────────────────────────────────────────────
 // TTS 引擎
 // ──────────────────────────────────────────────────────────────
@@ -1674,6 +1773,83 @@ async function apply(ctx) {
       },
     })))
 
+    // 4) send_image 工具 [2026-08-23]：agent 主动发图（仿 send_voice）
+    disposers.push(ctx.tools.register(defineTool({
+      name: 'send_image',
+      description: '向用户发送一张图片：把本地图片文件（png/jpg/jpeg/gif/webp）作为独立图片横条出现在聊天里（可点开放大、可翻查、手机可看）。'
+        + '【何时调用】① 用户明确要求"发张图/发个图片/把这张图发给我"；② 你判断发图比文字更直观时（截图、示意图、生成的图）；③ 用户让你"把刚才生成的/看到的图发出来"。'
+        + '【imagePath】本地图片文件的绝对路径；不支持远程 URL（先下载到本地再传）。'
+        + '【alt】可选的图片说明文字（显示在图片下方，也作为无障碍标签）。',
+      parameters: {
+        imagePath: {
+          type: 'string', required: true,
+          description: '要发送的本地图片文件绝对路径（png/jpg/jpeg/gif/webp）',
+        },
+        alt: {
+          type: 'string',
+          description: '可选的图片说明文字（显示在图片下方/作为无障碍标签）',
+        },
+      },
+      output: {
+        schema: {
+          type: 'object',
+          additionalProperties: false,
+          properties: {
+            ok: { type: 'boolean', required: true },
+            attachmentId: { type: 'string' },
+            width: { type: 'number' },
+            height: { type: 'number' },
+            error: { type: 'string' },
+          },
+        },
+        render(_args, value) {
+          if (value.ok) {
+            return [{ type: 'text', text: `图片已发送（attachmentId: ${value.attachmentId}，${value.width}×${value.height}）` }]
+          }
+          return [{ type: 'text', text: `图片发送失败：${value.error ?? '未知错误'}` }]
+        },
+      },
+      async execute(args, exec) {
+        const agent = exec?.agent
+        if (agent === undefined) return { ok: false, error: 'no session context (tool exec signature unsupported)' }
+        const session = agent.session
+        const imagePath = String(args.imagePath ?? '').trim()
+        if (imagePath === '') return { ok: false, error: 'imagePath is empty' }
+        const mediaType = sniffImageType(imagePath)
+        if (mediaType === undefined) return { ok: false, error: 'unsupported image type (png/jpg/jpeg/gif/webp expected)' }
+        const alt = typeof args.alt === 'string' ? args.alt.trim() : undefined
+        try {
+          const data = new Uint8Array(await readFile(imagePath))
+          const attachment = await saveImageFile(voiceStorageRoot(), data, mediaType)
+          let turn = 0
+          try {
+            turn = session.events
+              .filter((event) => event.type === 'turn/start')
+              .at(-1)?.data.turn ?? 0
+          } catch { /* rc.7 结构差异：忽略 */ }
+          try {
+            session.append('image/reply', {
+              turn,
+              attachmentId: attachment.attachmentId,
+              mediaType: attachment.mediaType,
+              bytes: attachment.bytes,
+              width: attachment.width,
+              height: attachment.height,
+              ...(alt === undefined || alt === '' ? {} : { alt }),
+            })
+          } catch { /* rc.7 无 append：忽略 */ }
+          return {
+            ok: true,
+            attachmentId: attachment.attachmentId,
+            width: attachment.width,
+            height: attachment.height,
+          }
+        } catch (error) {
+          return { ok: false, error: error instanceof Error ? error.message : 'unknown error' }
+        }
+      },
+    })))
+
     // 3.5) voice_config 实时查询工具 [2026-08-22]
     // send_voice 描述里的配置摘要是服务启动时的快照；AI 发送语音前可用本工具拿到最新配置
     disposers.push(ctx.tools.register(defineTool({

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@oadank/dsh-input-tools",
-  "version": "0.3.23",
+  "version": "0.3.24",
   "private": false,
   "type": "module",
   "main": "lib/index.js",