@goodandready/dsh-image-gen 0.10.0 → 0.10.2

package/lib/client.js

@@ -461,10 +461,12 @@ window.__ModuleLoader__.load({
     }
     const OUTPUT_FORMATS = ['png', 'jpeg', 'webp']
     const DELIVERY_MODES = ['link', 'image']
-    const PROVIDERS = ['fal', 'custom', 'codex', 'grok', 'local']
+    const PROVIDERS = ['fal', 'custom', 'codex', 'grok', 'local', 'seedream', 'gemini', 'replicate']
 
     /** Field specs in display order. */
+    const booleanField = (name) => ({ name, type: 'boolean', parse: (v) => v === 'true' || v === true, format: (v) => String(!!v) })
     const FIELDS = [
+      { field: 'enabled', spec: booleanField('enabled'), kind: 'select', options: ['true', 'false'], labelKey: 'f.enabled', hintKey: 'f.enabledHint' },
       { field: 'provider', spec: selectField('provider', PROVIDERS), kind: 'select', options: PROVIDERS, labelKey: 'f.provider', hintKey: 'f.providerHint', providerNeeds: { fal: 'ключ FAL', custom: 'адрес + ключ', codex: 'подписка ChatGPT', grok: 'подписка Grok', local: 'адрес сервера' } },
       { field: 'model', spec: textField('model'), kind: 'text', labelKey: 'f.model', hintKey: 'f.modelHint', placeholderKey: 'p.model', when: ['fal'] },
       { field: 'apiKeyEnv', spec: textField('apiKeyEnv'), kind: 'text', labelKey: 'f.apiKeyEnv', hintKey: 'f.apiKeyEnvHint', placeholderKey: 'p.apiKeyEnv', when: ['fal'] },
@@ -808,17 +810,38 @@ window.__ModuleLoader__.load({
       }
     }
 
-    function FalImageCard(props) {
+function FalImageCard(props) {
       const block = props.block
       const running = !('kind' in (block || {}))
       const failed = block && (block.isError || block.error !== undefined)
       const parsed = readResult(block)
-      const [editing, setEditing] = react.useState(false)
 
-      const [inpaintOpen, setInpaintOpen] = react.useState(false)
+      let args = {}
+      try {
+        const raw = (block && (block.call ? block.call.argsRaw : block.argsRaw)) || ''
+        if (raw) args = JSON.parse(raw)
+      } catch (e) { /* ignore */ }
 
+      let prompt = ''
+      try {
+        const raw = (block && (block.call ? block.call.argsRaw : block.argsRaw)) || ''
+        prompt = raw ? (JSON.parse(raw).prompt || '') : ''
+      } catch (e) { prompt = '' }
+
+      const repeatText = (() => {
+        const parts = [String(args.prompt || '').trim()]
+        const size = args.image_size || args.size
+        const seed = args.seed
+        if (size) parts.push(`Размер: ${size}`)
+        if (seed !== undefined) parts.push(`Сид: ${seed}`)
+        return parts.filter(Boolean).join(', ')
+      })()
+
+      const [editing, setEditing] = react.useState(false)
+      const [inpaintOpen, setInpaintOpen] = react.useState(false)
       const [compareSlider, setCompareSlider] = react.useState(50)
       const [showCompare, setShowCompare] = react.useState(false)
+      const [editText, setEditText] = react.useState('')
 
       const onEditPrompt = () => {
         const textToInsert = prompt || repeatText || ''
@@ -883,12 +906,7 @@ window.__ModuleLoader__.load({
         sendActionPrompt(`Удали фон у этого изображения: remove_background(image: "${ref}")`)
       }
 
-      const [editText, setEditText] = react.useState('')
-      let prompt = ''
-      try {
-        const raw = (block && (block.call ? block.call.argsRaw : block.argsRaw)) || ''
-        prompt = raw ? (JSON.parse(raw).prompt || '') : ''
-      } catch (e) { prompt = '' }
+
 
       const head = react.createElement('div', { className: 'fal-head' },
         running ? 'Генерирую изображение…' : failed ? 'Генерация не удалась' : 'Изображение')
@@ -897,19 +915,7 @@ window.__ModuleLoader__.load({
       // промптом и параметрами, извлечёнными из исходного вызова инструмента.
       // Дальше модель сама решает, как сгенерировать, — это просто повторный
       // запрос, а не вызов инструмента из карточки.
-      let args = {}
-      try {
-        const raw = (block && (block.call ? block.call.argsRaw : block.argsRaw)) || ''
-        if (raw) args = JSON.parse(raw)
-      } catch (e) { /* ignore */ }
-      const repeatText = (() => {
-        const parts = [String(args.prompt || '').trim()]
-        const size = args.image_size || args.size
-        const seed = args.seed
-        if (size) parts.push(`Размер: ${size}`)
-        if (seed !== undefined) parts.push(`Сид: ${seed}`)
-        return parts.filter(Boolean).join(', ')
-      })()
+
       const repeat = async () => {
         try {
           const sessions = props.sessions

package/lib/index.js

@@ -16,10 +16,14 @@ import z from '@deepseek-ai/schemastery'
 import { defineTool } from '@deepseek-ai/dsh-tools'
 import { credentialRef } from '@deepseek-ai/dsh-credentials'
 import { mkdir, readFile, writeFile } from 'node:fs/promises'
+import { existsSync, unlinkSync, accessSync } from 'node:fs'
 import os from 'node:os'
 import path from 'node:path'
 import {
   computeGenerationHash,
+  clampProviderCount,
+  buildEndpointUrl,
+  createAbortError,
   IMAGE_SIZES,
   OUTPUT_FORMATS,
   PROVIDER_KEYS,
@@ -54,9 +58,13 @@ const NS = 'dsh-image-gen'
 // настройки лежат под старым именем. Оно читается как запасное и переносится
 // под новое имя один раз — молча, при первом запуске после обновления.
 const LEGACY_NS = 'dsh-fal-image-gen'
-export const inject = ['tools', 'attachments', 'credentials', 'webServer', 'settings', 'llm']
+export const inject = ['tools', 'attachments', 'credentials', 'webServer', 'settings', 'llm', 'systemPrompt']
 
 export const Config = z.object({
+  enabled: z
+    .boolean()
+    .description('Master switch for image generation. When false, tools refuse to execute.')
+    .default(true),
   provider: z
     .string()
     .description(`Which provider generates the image. One of: ${PROVIDER_KEYS.join(', ')}. `
@@ -68,6 +76,7 @@ export const Config = z.object({
     .default('fal-ai/flux-2/klein/9b'),
   apiKeyEnv: z
     .string()
+    .role('credential-ref')
     .description('Credential reference / env var holding the FAL API key (the "Key " auth prefix is added automatically when missing).')
     .default('FAL_API_KEY'),
   baseURL: z
@@ -109,6 +118,7 @@ export const Config = z.object({
     .default(''),
   customKeyEnv: z
     .string()
+    .role('credential-ref')
     .description('provider=custom: credential reference / env var holding the API key. '
       + 'Empty means no authorization header, for gateways that need none.')
     .default('OPENAI_API_KEY'),
@@ -119,6 +129,7 @@ export const Config = z.object({
     .default('black-forest-labs/flux-schnell'),
   replicateKeyEnv: z
     .string()
+    .role('credential-ref')
     .description('provider=replicate: credential reference / env var holding API token.')
     .default('REPLICATE_API_TOKEN'),
 
@@ -154,6 +165,7 @@ export const Config = z.object({
     .default(7),
   seedreamKeyEnv: z
     .string()
+    .role('credential-ref')
     .description('provider=seedream: credential reference / env var holding the API key.')
     .default('SEEDREAM_API_KEY'),
   seedreamModel: z
@@ -162,6 +174,7 @@ export const Config = z.object({
     .default('seedream-4.0'),
   geminiKeyEnv: z
     .string()
+    .role('credential-ref')
     .description('provider=gemini: credential reference / env var holding the Google API key.')
     .default('GEMINI_API_KEY'),
   geminiModel: z
@@ -217,7 +230,12 @@ export async function resolveSource(ctx, exec, ref) {
   }
   const p = path.resolve(sessionCwd || process.cwd(), ref)
   const bytes = await readFile(p)
-  return { bytes, mediaType: 'image/png' }
+  const ext = path.extname(p).toLowerCase()
+  const mediaType = ext === '.jpg' || ext === '.jpeg' ? 'image/jpeg'
+    : ext === '.webp' ? 'image/webp'
+    : ext === '.svg' ? 'image/svg+xml'
+    : 'image/png'
+  return { bytes, mediaType }
 }
 
 export function slugify(input) {
@@ -295,7 +313,7 @@ export async function findCached(entries, seed, prompt) {
 /** Вернуть запись из кэша, если файл существует; иначе undefined. */
 export async function cachedResult(entry) {
   if (!entry || !entry.path) return undefined
-  try { require('node:fs').accessSync(entry.path) } catch (e) { return undefined }
+  if (!entry.path || !existsSync(entry.path)) return undefined
   return {
     path: entry.path,
     url: entry.attachmentId ? `/dsh-image-gen/image?id=${encodeURIComponent(entry.attachmentId)}` : '',
@@ -326,8 +344,8 @@ export async function pruneHistory(entries, pruneDays) {
   for (const e of entries) {
     const created = e.createdAt ? Date.parse(e.createdAt) : NaN
     if (Number.isFinite(created) && created < cutoff) {
-      try { require('node:fs').unlinkSync(e.path) } catch (err) { /* файл уже удалён */ }
-      try { require('node:fs').unlinkSync(e.path.replace(/\.[^.]+$/, '.json')) } catch (err) { /* sidecar */ }
+      try { unlinkSync(e.path) } catch (err) { /* файл уже удалён */ }
+      try { unlinkSync(e.path.replace(/\.[^.]+$/, '.json')) } catch (err) { /* sidecar */ }
       continue
     }
     kept.push(e)
@@ -408,6 +426,17 @@ export async function enhancePrompt(ctx, cfg, prompt, signal, provider) {
 }
 
 export function apply(ctx, config) {
+  if (ctx.systemPrompt && typeof ctx.systemPrompt.section === 'function') {
+    ctx.systemPrompt.section({
+      name: 'tool:image-generation',
+      order: 121,
+      text: 'To create images, call generate_image with a prompt and optional count, aspect_ratio, quality, or output_dir. '
+        + 'The tool writes images to disk and shows them in the conversation. The model cannot see the image directly in the text stream; '
+        + 'for quality inspection, verification of expected elements, or autonomous correction (Vision Loop), call analyze_image from the @goodandready/dsh-vision-bridge plugin '
+        + 'or inspect_image_quality on the returned path.',
+    })
+  }
+
 
   // The Web card edits this namespace; without registering it the card binds to
   // a namespace nobody declared, stays unready and renders nothing — which is
@@ -494,7 +523,7 @@ export function apply(ctx, config) {
         res.end(JSON.stringify({ error: 'GET only' }))
         return
       }
-      const exists = (p) => { try { return require('node:fs').existsSync(p) } catch { return false } }
+      const exists = (p) => existsSync(p)
       const entries = await readHistory()
       const withThumbs = filterHistory(entries, exists).map((e) => ({
         ...e,
@@ -526,7 +555,7 @@ export function apply(ctx, config) {
         },
         aspect_ratio: {
           type: 'string',
-          enum: ['1:1', '16:9', '9:16', '4:3', '3:4'],
+          enum: ['1:1', '16:9', '9:16', '4:3', '3:4', '3:2', '2:3'],
           description: 'Optional aspect ratio; overrides image_size when both are given. Sent as width/height to providers that accept pixels.',
         },
         seed: {
@@ -542,6 +571,10 @@ export function apply(ctx, config) {
           type: 'string',
           description: 'Optional file name stem for the saved image (defaults to a slug of the prompt).',
         },
+        output_dir: {
+          type: 'string',
+          description: 'Optional custom directory to write images into (relative to session cwd or absolute). Defaults to outputDir from settings.',
+        },
         count: {
           type: 'integer',
           description: 'Number of variations to generate in one call, 1-4 (default 1). Cost scales with count.',
@@ -566,7 +599,7 @@ export function apply(ctx, config) {
         },
         quality: {
           type: 'string',
-          enum: ['low', 'medium', 'high'],
+          enum: ['auto', 'low', 'medium', 'high'],
           description: 'Optional quality for providers that support it (seedream, gemini). Ignored otherwise.',
         },
         style: {
@@ -589,7 +622,7 @@ export function apply(ctx, config) {
       output: {
         schema: {
           type: 'object',
-          additionalProperties: false,
+          additionalProperties: true,
           properties: {
             path: { type: 'string' },
             url: { type: 'string' },
@@ -630,7 +663,7 @@ export function apply(ctx, config) {
         },
       render(args, value) {
           const list = value.images || [value]
-          const lines = list.map((img) => `Image generated (${img.width}×${img.height}, ${img.format}, seed ${img.seed}): ${img.path}`)
+          const lines = list.map((img) => `Generated image (${img.width}×${img.height}, ${img.format}, seed ${img.seed}) via ${value.provider || 'image-gen'}/${value.model || 'default'}: ${img.path}`)
           const summary = lines.join('\n')
           if (live().deliverAs !== 'image') {
             const urls = list.map((img) => img.url).filter(Boolean)
@@ -645,6 +678,9 @@ export function apply(ctx, config) {
       timeoutMs: config.timeoutMs + 30000,
       async execute(args, exec) {
         const cfg = live()
+        if (cfg.enabled === false) {
+          throw new Error('Image generation is disabled in settings (dsh-image-gen.enabled is false); enable it in Settings → Image generation.')
+        }
         const provider = PROVIDER_KEYS.includes(cfg.provider) ? cfg.provider : 'fal'
         const deliverAs = cfg.deliverAs
         const size = args.aspect_ratio ? 'custom' : (args.image_size ?? cfg.defaultSize)
@@ -664,12 +700,14 @@ export function apply(ctx, config) {
         const source = args.source_image ? await resolveSource(ctx, exec, args.source_image) : undefined
         const mask = args.mask ? await resolveSource(ctx, exec, args.mask) : undefined
         const enhanced = await enhancePrompt(ctx, cfg, args.prompt, exec.signal, provider)
-        const styleSuffix = cfg.stylePreset ? `, ${cfg.stylePreset}` : ''
-        const styledPrompt = args.style ? applyStylePreset(enhanced.prompt, args.style) : enhanced.prompt;
-        const effectivePrompt = styledPrompt + styleSuffix
+        const effectiveStyle = args.style || cfg.stylePreset
+        const styleInfo = resolveStylePreset(effectiveStyle, args.negative_prompt, args.guidance_scale)
+        const effectivePrompt = styleInfo.promptSuffix ? `${enhanced.prompt}, ${styleInfo.promptSuffix}` : enhanced.prompt
+        const effectiveNegative = styleInfo.negativePrompt
+        const effectiveGuidance = styleInfo.guidanceScale
         const providers = makeProviders(
           { fetchImpl: fetch, resolveKey: (ref) => resolveApiKey(ctx, ref), cfg, subscriptionImages },
-          { prompt: effectivePrompt, size, format, seed: args.seed, signal: exec.signal, negativePrompt: args.negative_prompt, guidanceScale: args.guidance_scale, source, mask, strength: args.strength, quality: args.quality, style: args.style, aspectPixels },
+          { prompt: effectivePrompt, size, format, seed: args.seed, signal: exec.signal, negativePrompt: effectiveNegative, guidanceScale: effectiveGuidance, source, mask, strength: args.strength, quality: args.quality, style: args.style, aspectPixels },
         )
         // Подписочные провайдеры (codex/grok) отдают {ok:false, reason} вместо исключения:
         // отказ должен дойти до модели текстом. Без проверки execute шёл дальше с пустыми
@@ -703,7 +741,8 @@ export function apply(ctx, config) {
           })
 
           const sessionCwd = exec.agent?.session?.header?.cwd
-          const outDir = path.resolve(sessionCwd || process.cwd(), cfg.outputDir)
+          const targetDir = (args.output_dir && String(args.output_dir).trim()) || cfg.outputDir || 'generated/images'
+          const outDir = path.resolve(sessionCwd || process.cwd(), targetDir)
           await mkdir(outDir, { recursive: true })
           const filePath = path.join(outDir, name)
           await writeFile(filePath, bytes)
@@ -734,13 +773,28 @@ export function apply(ctx, config) {
             }), null, 2),
           )
 
+          const cacheHash = computeGenerationHash({
+            provider,
+            model: cfg.model,
+            prompt: promptArg,
+            seed: gen.seed,
+            size: args.image_size || cfg.defaultSize,
+            style: args.style || cfg.stylePreset,
+          })
           const entry = {
             path: filePath,
             prompt: promptArg,
             provider,
+            model: cfg.model,
             size,
             format,
             seed: gen.seed,
+            cacheHash,
+            thumbnailUrl: localUrl,
+            width: gen.width || attachment.width,
+            height: gen.height || attachment.height,
+            mediaType,
+            bytes: attachment.bytes,
             createdAt: new Date().toISOString(),
             attachmentId: attachment.attachmentId,
           }
@@ -817,7 +871,7 @@ export function apply(ctx, config) {
           }
         }
         const first = images[0]
-        return { ...first, images }
+        return { channel: provider, provider, model: cfg.model || cfg.customModel, ...first, images }
       },
     }),
   )

package/lib/providers.js

@@ -1,4 +1,33 @@
 
+export const PROVIDER_MAX_COUNTS = {
+  fal: 4,
+  replicate: 4,
+  custom: 10,
+  seedream: 10,
+  gemini: 4,
+  codex: 1,
+  grok: 1,
+  local: 4,
+}
+
+export function clampProviderCount(provider, count) {
+  const max = PROVIDER_MAX_COUNTS[provider] || 4
+  return normalizeCount(count, max)
+}
+
+export function buildEndpointUrl(baseURL, pathSuffix) {
+  const base = String(baseURL || '').trim().replace(/\/+$/, '')
+  const suffix = String(pathSuffix || '').trim().replace(/^\/+/, '')
+  return `${base}/${suffix}`
+}
+
+export function createAbortError(message = 'Image generation cancelled') {
+  const err = new Error(message)
+  err.name = 'AbortError'
+  return err
+}
+
+
 /** Fast Laplacian-like sharpness and entropy estimator across image scanlines. */
 export function estimateSharpnessAndVariance(bytes) {
   const buf = Buffer.isBuffer(bytes) ? bytes : Buffer.from(bytes || [])
@@ -132,10 +161,10 @@ export async function tryGenerate(generators, order, seed, promptArg) {
   throw new Error(`Image generation failed on all providers. ${refusals.join('; ')}`)
 }
 
-export function normalizeCount(count) {
+export function normalizeCount(count, max = 4) {
   const n = Math.floor(Number(count))
   if (!Number.isFinite(n)) return 1
-  return Math.max(1, Math.min(4, n))
+  return Math.max(1, Math.min(max, n))
 }
 
 // Размеры, которыми оперируют подписки, — свой набор, не похожий на именованные
@@ -238,7 +267,7 @@ export async function pollStatus(fetchImpl, statusUrl, key, signal, pollInterval
         }, { once: true })
       }
     })
-    if (signal.aborted) throw new Error('FAL generation cancelled')
+    if (signal?.aborted) throw new Error('FAL generation cancelled')
     const res = await fetchImpl(statusUrl, { headers: { Authorization: falAuthHeader(key) }, signal })
     const data = await res.json().catch(() => ({}))
     const status = data.status
@@ -268,6 +297,8 @@ export const ASPECT_RATIOS = {
   '9:16': [768, 1344],
   '4:3': [1152, 896],
   '3:4': [896, 1152],
+  '3:2': [1152, 768],
+  '2:3': [768, 1152],
 }
 
 /** Размер в пикселях: aspectPixels (если задан) или из именованного размера. */
@@ -370,7 +401,7 @@ export function makeProviders(deps, job) {
 
     // response_format не отправляем: новые модели OpenAI его отвергают, а ответ
     // всё равно приходит либо base64, либо ссылкой — принимаем оба.
-    const endpoint = source ? `${base}/images/edits` : `${base}/images/generations`
+    const endpoint = buildEndpointUrl(base, source ? 'images/edits' : 'images/generations')
     const res = await fetchImpl(endpoint, {
       method: 'POST',
       headers,
@@ -383,11 +414,18 @@ export function makeProviders(deps, job) {
             size: cfg.customSize || (aspectPixels ? aspectPixels.join('x') : SIZE_PIXELS[size]) || size,
             ...(negativePrompt !== undefined ? { negative_prompt: negativePrompt } : {}),
             ...(guidanceScale !== undefined ? { guidance_scale: guidanceScale } : {}),
+            ...(quality && quality !== 'auto' ? { quality } : {}),
           }),
       signal,
     })
     const data = await res.json().catch(() => ({}))
     if (!res.ok) {
+      if (res.status === 503) {
+        const msg = (data?.error?.message || JSON.stringify(data)).toLowerCase()
+        if (msg.includes('no available channel') || msg.includes('model_not_found')) {
+          throw new Error(`Custom image API has no provisioned channel for "${cfg.customModel}" (HTTP 503); check gateway channel and routing configuration`)
+        }
+      }
       const detail = data?.error?.message || JSON.stringify(data).slice(0, 600)
       throw new Error(`Image API failed (HTTP ${res.status}): ${detail}`)
     }
@@ -540,7 +578,7 @@ export function makeProviders(deps, job) {
     const deadline = Date.now() + cfg.timeoutMs
     let attempt = 0
     for (;;) {
-      if (signal.aborted) throw new Error('Local ComfyUI generation cancelled')
+      if (signal?.aborted) throw new Error('Local ComfyUI generation cancelled')
       if (Date.now() > deadline) throw new Error(`Local ComfyUI timed out after ${cfg.timeoutMs} ms`)
       const cDelay = calculateBackoff(attempt++, cfg.pollIntervalMs || 1000, 5000)
       await new Promise((resolve) => {
@@ -663,7 +701,7 @@ export function makeProviders(deps, job) {
       const deadline = Date.now() + (cfg.timeoutMs || 180000)
       let attempt = 0
     while (pred.status !== 'succeeded') {
-        if (signal.aborted) throw new Error('Replicate generation cancelled')
+        if (signal?.aborted) throw new Error('Replicate generation cancelled')
         if (Date.now() > deadline) throw new Error('Replicate generation timed out')
         if (pred.status === 'failed' || pred.status === 'canceled') {
           throw new Error(`Replicate failed: ${pred.error || 'unknown error'}`)

package/package.json

@@ -1,6 +1,6 @@
 {
   "name": "@goodandready/dsh-image-gen",
-  "version": "0.10.0",
+  "version": "0.10.2",
   "description": "Image generation for DeepSeek Harness: a generate_image tool with pluggable providers — the FAL queue, any OpenAI-compatible images API, or a ChatGPT/Grok subscription with no API key at all. The picture is shown inline in the conversation; the model receives either a link (works with any chat model) or the image itself (needs dsh-vision-bridge or a vision-capable model).",
   "keywords": [
     "deepseek-harness",