mirror of https://git.openapi.site/https://github.com/desirecore/market.git synced 2026-07-23 06:23:42 +08:00

Files

xyx d83ab2d842 feat: 上架 minimax-video-gen（移除 listed: false） (#35 )

## Summary

- 移除 `minimax-video-gen` 的 `market.listed: false`，使其在市场页面可见，用户可手动安装

Remove `market.listed: false` from `minimax-video-gen` so it appears in
the market UI for manual installation.

## Changes

- **skills/minimax-video-gen/SKILL.md**: remove `listed: false` (1 line
deletion)

2026-07-02 19:03:14 +08:00

9.5 KiB

Raw Blame History

name, description, license, version, type, risk_level, status, disable-model-invocation, provider, tags, requires, metadata, market

name

description

license

version

type

risk_level

status

disable-model-invocation

provider

minimax-video-gen Skill

Mandatory Rules (violation will cause failure)

Must use HTTPS to access agent-service — https://127.0.0.1:${PORT} with -k to skip certificate verification
Use Bash curl throughout — do not use the HttpRequest tool or Python
Polling interval is 10 seconds — use sleep 10 to wait
Must use providerId — do not use provider, or requests will route to the wrong upstream

Full Execution Flow

Prerequisites

The user has already configured and enabled a Provider with video_gen service in Resource Manager → Compute and entered the API Key
agent-service is running

Core Concept: Four-Step Asynchronous Flow

Video generation uses an asynchronous task model via the NewAPI gateway:

Submit task: POST /video/generations to create a task, receive a task_id
Poll status: GET /videos/{task_id} until status is "completed"
Download & upload: download from data.metadata.url, upload to media-store
Display: use dc-media:// protocol to show the video

Provider Selection (Important)

Use providerId (not provider) to target the correct gateway. When the provider is a NewAPI gateway (e.g. desirecore-cloud), use providerId. Do not use provider: "minimax" — this would route to the MiniMax native host, which does not support the NewAPI /video/generations path.

{
  "providerId": "desirecore-cloud",
  "serviceType": "video_gen",
  "endpoint": "/video/generations",
  ...
}

Model Selection and Fallback Strategy

Model	Supported Modes	Characteristics	Use Case
MiniMax-Hailuo-2.3	Text-to-video + image-to-video	Highest quality, default first choice	Default when the user doesn't specify
MiniMax-Hailuo-2.3-fast	Image-to-video only	Fast, 50% lower cost	Fallback when quota is insufficient in image-to-video scenarios

Fallback rules (mandatory):

Use MiniMax-Hailuo-2.3 by default
When text-to-video (T2V) quota is insufficient: MiniMax-Hailuo-2.3-fast does not support text-to-video and cannot be used as a fallback. Inform the user directly that the quota is insufficient and suggest waiting for the quota to reset or switching to another video generation service (such as Kling)
When image-to-video (I2V) quota is insufficient: fall back to MiniMax-Hailuo-2.3-fast and inform the user "switched to the fast model for generation"
If the user is doing image-to-video and explicitly requests fast generation, use MiniMax-Hailuo-2.3-fast directly

Step 1: Submit a Text-to-Video Task

Important: The endpoint is /video/generations (with an "s"), NOT /video_generation.

PORT=$(cat ${DESIRECORE_ROOT}/agent-service.port)
curl -sk -X POST "https://127.0.0.1:${PORT}/api/media-proxy" \
  -H "Content-Type: application/json" \
  -d '{
    "providerId": "desirecore-cloud",
    "serviceType": "video_gen",
    "endpoint": "/video/generations",
    "body": {
      "model": "MiniMax-Hailuo-2.3",
      "prompt": "Video content described by the user",
      "size": "768P",
      "duration": 6
    },
    "responseType": "json"
  }'

Optional parameters (add to the body):

"duration": video length in seconds (6 or 10)
"size": "768P" or "1080P" (use size, not resolution)

Extract data.task_id from the JSON response.

Successful response example:

{
  "success": true,
  "data": {
    "id": "task_xxx",
    "task_id": "task_xxx",
    "object": "video",
    "model": "MiniMax-Hailuo-2.3",
    "status": "queued",
    "progress": 0,
    "created_at": 1780995870
  },
  "statusCode": 200
}

Step 1 (alternative): Image-to-Video

PORT=$(cat ${DESIRECORE_ROOT}/agent-service.port)
curl -sk -X POST "https://127.0.0.1:${PORT}/api/media-proxy" \
  -H "Content-Type: application/json" \
  -d '{
    "providerId": "desirecore-cloud",
    "serviceType": "video_gen",
    "endpoint": "/video/generations",
    "body": {
      "model": "MiniMax-Hailuo-2.3",
      "prompt": "Describe the dynamic changes of the scene in the image",
      "first_frame_image": "https://image-URL",
      "size": "768P"
    },
    "responseType": "json"
  }'

Step 2: Poll the Task Status

Call once every 10 seconds until status is "completed" or "failed". Replace TASK_ID with the task_id returned in Step 1.

PORT=$(cat ${DESIRECORE_ROOT}/agent-service.port)
TASK_ID="task_id returned from step 1"
curl -sk -X POST "https://127.0.0.1:${PORT}/api/media-proxy" \
  -H "Content-Type: application/json" \
  -d "{
    \"providerId\": \"desirecore-cloud\",
    \"serviceType\": \"video_gen\",
    \"endpoint\": \"/videos/${TASK_ID}\",
    \"method\": \"GET\",
    \"responseType\": \"json\"
  }"

In-progress statuses may be: "queued" or "in_progress".

Polling response (in progress):

{
  "success": true,
  "data": {
    "id": "task_xxx",
    "status": "in_progress",
    "progress": 42
  }
}

Polling response (completed):

{
  "success": true,
  "data": {
    "id": "task_xxx",
    "status": "completed",
    "progress": 100,
    "metadata": {
      "url": "https://.../output_aigc.mp4?..."
    }
  }
}

Step 3: Download and Upload to media-store

Extract the video download URL from data.metadata.url in the completed Step 2 response. The URL is valid for 24 hours; download immediately.

PORT=$(cat ${DESIRECORE_ROOT}/agent-service.port)
VIDEO_URL="data.metadata.url from step 2"
curl -sL "$VIDEO_URL" -o /tmp/minimax-video.mp4 && \
curl -sk -X POST "https://127.0.0.1:${PORT}/api/media/upload" \
  -F "file=@/tmp/minimax-video.mp4;type=video/mp4"

Extract the mediaId field from the JSON response.

Step 4: Display the Video Using the dc-media Protocol

Write Markdown image syntax directly in your reply (the frontend will automatically recognize the video extension and render a player):

![Video description](dc-media://replace-with-mediaId)

Error Handling

status: "failed": video generation failed; explain to the user
success: false + error: "No matching provider found": No enabled provider with video_gen service found
success: false + error: "API Key not configured": API Key has not been entered
Insufficient quota (errors related to statusCode: 429, insufficient_quota, balance): text-to-video cannot fall back (the Fast model does not support T2V); inform the user of insufficient quota; image-to-video can switch to MiniMax-Hailuo-2.3-fast and retry from Step 1
Polling exceeds 10 minutes without completion: inform the user that the task may have timed out

Notes

Video generation is asynchronous and typically takes 2–10 minutes
A polling interval of 10 seconds is recommended
The download URL is valid for 24 hours
If the user does not explicitly request otherwise, by default pass "size": "768P" and "duration": 6
Do not proxy the download URL through media-proxy — use curl -sL "$VIDEO_URL" to download directly

9.5 KiB Raw Blame History Unescape Escape