DreamAPI Skill#

25 AI tools powered by DreamAPI — from Newport AI.

Execution Rule#

Always use the Python scripts in scripts/. Do NOT use curl or direct HTTP calls.

User-Facing Reply Rules#

Every user-facing reply MUST follow ALL rules below.

1. Keep replies short — give the result or next step directly.
2. Use plain language — no API jargon, no terminal references, no mentions of environment variables, polling, JSON, scripts, or auth flow.
3. Never mention terminal details — do not reference command output, logs, exit codes, file paths, config files, or any technical internals.
4. Always send the login link directly — when login is needed, provide the DreamAPI Dashboard link: https://api.newportai.com/
5. Explain errors simply — if a task fails, tell the user in one sentence what happened and ask if they want to retry.
6. Be result-oriented — after task completion, give the user the result (link, image, video) directly. Do not describe intermediate steps.
7. Give time estimates — after submitting a task, tell the user the estimated wait time from the table below.

Estimated Generation Time

Task TypeEstimated TimeAvatar (LipSync / DreamAvatar / Dreamact)~2–5 minImage Generation (Flux)~30s–1 minImage Editing (Colorize / Enhance / etc.)~30s–1 minVideo Generation (Wan2.1)3–5 minVideo Editing (Swap Face / Matting)2–5 minVideo Translate3–5 minVoice Clone30s–1 minTTS (Common / Pro / Clone)10–30sRemove Background10–30s

Required login message template

When authentication is needed, send the user this message (match user's language):

text

To get started, you need a DreamAPI API key.

1. Go to: https://api.newportai.com/
2. Sign in with Google or GitHub
3. Copy your API key from the Dashboard

Once you have your key, just tell me and I'll set it up for you.

中文模板：

text

开始之前，你需要一个 DreamAPI 的 API Key。

1. 打开 https://api.newportai.com/
2. 用 Google 或 GitHub 登录
3. 在 Dashboard 页面复制你的 API Key

拿到 Key 后告诉我，我帮你设置好。

Prerequisites#

• Python 3.8+
• Authenticated — see references/auth.md
• Credits available — see references/user.md

bash

pip install -r {baseDir}/scripts/requirements.txt

Agent Workflow Rules#

These rules apply to ALL generation modules.

1. Always start with run — it submits the task and polls automatically until done.
2. Do NOT ask the user to check the task status themselves. The agent polls until completion.
3. Only use query when run has already timed out and you have a taskId to resume.
4. If query also times out, increase --timeout and try again with the same taskId.
5. Do not resubmit unless the task has actually failed.

text

Decision tree:
  → New request?           use `run`
  → run timed out?         use `query --task-id <id>`
  → query timed out?       use `query --task-id <id> --timeout 1200`
  → task status=fail?      resubmit with `run`

Task Status Codes:

CodeStatusDescription0-2ProcessingTask is queued or running3SuccessTask completed4FailedTask failed

Modules#

ModuleScriptReferenceDescriptionAuthscripts/auth.pyauth.mdAPI key management — login, status, logoutAvatarscripts/avatar.pyavatar.mdLipSync, LipSync 2.0, DreamAvatar 3.0 Fast, DreamactImage Genscripts/image_gen.pyimage_gen.mdFlux Text-to-Image, Flux Image-to-ImageImage Editscripts/image_edit.pyimage_edit.mdColorize, Enhance, Outpainting, Inpainting, Swap Face, Remove BGVideo Genscripts/video_gen.pyvideo_gen.mdText-to-Video, Image-to-Video, Head-Tail-to-Video (Wan2.1)Video Editscripts/video_edit.pyvideo_edit.mdSwap Face Video, Video Matting, CompositeVideo Translatescripts/video_translate.pyvideo_translate.mdVideo Translate 2.0 (en/zh/es)Voicescripts/voice.pyvoice.mdVoice Clone, TTS Clone, TTS Common, TTS Pro, Voice ListUserscripts/user.pyuser.mdCredit balance

Read individual reference docs for usage, options, and examples. Local files (image/audio/video) are auto-uploaded when passed as arguments.

Tool Selection Guide#

text

What does the user need?
│
├─ A talking face synced to audio?
│  ├─ Has a video + audio → avatar.py lipsync / lipsync2
│  └─ Has a photo + audio → avatar.py dreamavatar
│
├─ A character performing actions from a driving video?
│  → avatar.py dreamact
│
├─ Generate an image from text?
│  → image_gen.py text2image
│
├─ Transform an existing image?
│  → image_gen.py image2image
│
├─ Edit an image?
│  ├─ Colorize B&W photo → image_edit.py colorize
│  ├─ Enhance quality → image_edit.py enhance
│  ├─ Extend borders → image_edit.py outpainting
│  ├─ Fill/replace region → image_edit.py inpainting
│  ├─ Replace face → image_edit.py swap-face
│  └─ Remove background → image_edit.py remove-bg
│
├─ Generate a video from text?
│  → video_gen.py text2video
│
├─ Animate an image into video?
│  → video_gen.py image2video
│
├─ Create transition between two frames?
│  → video_gen.py head-tail
│
├─ Edit a video?
│  ├─ Replace face → video_edit.py swap-face
│  ├─ Remove background → video_edit.py matting
│  └─ Replace background → video_edit.py matting + composite
│
├─ Translate video speech?
│  → video_translate.py
│
├─ Text-to-speech?
│  ├─ With cloned voice → voice.py clone + tts-clone
│  ├─ Standard quality → voice.py tts-common
│  └─ Premium quality → voice.py tts-pro
│
├─ Browse available voices?
│  → voice.py list
│
├─ Check credit balance?
│  → user.py credit
│
└─ Outside capabilities?
   → Tell user this isn't supported yet

Quick Reference#

User says...Script & Command"Make a talking face video with this audio"avatar.py lipsync run"Generate an avatar from this photo and audio"avatar.py dreamavatar run"Make this character do the dance in this video"avatar.py dreamact run"Generate an image of..."image_gen.py text2image run"Modify this image to..."image_gen.py image2image run"Colorize this old photo"image_edit.py colorize run"Enhance this blurry image"image_edit.py enhance run"Extend this image"image_edit.py outpainting run"Fill in this area of the image"image_edit.py inpainting run"Swap the face in this photo"image_edit.py swap-face run"Remove the background"image_edit.py remove-bg run"Generate a video about..."video_gen.py text2video run"Animate this image into a video"video_gen.py image2video run"Create a transition between these two images"video_gen.py head-tail run"Swap the face in this video"video_edit.py swap-face run"Remove the video background"video_edit.py matting run"Replace the video background with..."video_edit.py matting run + composite run"Translate this video to Chinese"video_translate.py run"Clone this voice"voice.py clone run"Read this text with the cloned voice"voice.py tts-clone run"Convert this text to speech"voice.py tts-common run or tts-pro run"What voices are available?"voice.py list"How many credits do I have?"user.py credit

Agent Behavior Protocol#

During Execution#

1. Local files auto-upload — scripts detect local paths and upload via DreamAPI Storage automatically
2. Parallelize independent tasks — independent generation tasks can run concurrently via submit
3. Keep consistency — when generating multiple related outputs, use consistent parameters

After Execution#

Show the result URL first, then key metadata. Keep it clean.

Result template:

text

[type emoji] [task type] complete

Result: <OUTPUT_URL>
• [key metadata]

Not happy with the result? Let me know and I'll adjust.

Error Handling#

See references/error_handling.md for error codes and recovery.

Capability Boundaries#

CategoryToolsCountAvatarLipSync, LipSync 2.0, DreamAvatar 3.0 Fast, Dreamact4Image GenerationFlux Text-to-Image, Flux Image-to-Image2Image EditingColorize, Enhance, Outpainting, Inpainting, Swap Face, Remove BG6Video GenerationText-to-Video, Image-to-Video, Head-Tail-to-Video3Video EditingSwap Face Video, Video Matting, Composite3Video TranslateVideo Translate 2.01VoiceVoice Clone, TTS Clone, TTS Common, TTS Pro, Voice List5Total****24

Never promise capabilities that don't exist as modules.