@acedatacloud/skills 2026.806.0 → 2026.807.0
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
package/package.json
CHANGED
|
@@ -1,6 +1,6 @@
|
|
|
1
1
|
{
|
|
2
2
|
"name": "@acedatacloud/skills",
|
|
3
|
-
"version": "2026.
|
|
3
|
+
"version": "2026.807.0",
|
|
4
4
|
"description": "Agent Skills for AceDataCloud AI services — music, image, video generation, LLM chat, web search. Compatible with Claude Code, GitHub Copilot, Gemini CLI, OpenAI Codex, and 30+ AI coding agents.",
|
|
5
5
|
"keywords": [
|
|
6
6
|
"agent-skills",
|
|
@@ -0,0 +1,100 @@
|
|
|
1
|
+
---
|
|
2
|
+
name: minimax-video
|
|
3
|
+
description: Generate MiniMax H3 videos from text, up to nine reference images, or up to three audio references through AceDataCloud. Use for text-to-video, multi-image video, audio-guided video, and MiniMax H3 task polling.
|
|
4
|
+
license: Apache-2.0
|
|
5
|
+
metadata:
|
|
6
|
+
author: acedatacloud
|
|
7
|
+
version: "1.0"
|
|
8
|
+
compatibility: Requires ACEDATACLOUD_API_TOKEN in .env (see _shared/authentication.md).
|
|
9
|
+
---
|
|
10
|
+
|
|
11
|
+
# MiniMax H3 Video Generation
|
|
12
|
+
|
|
13
|
+
Generate 4–15 second videos through `POST https://api.acedata.cloud/minimax/videos`.
|
|
14
|
+
|
|
15
|
+
> **Setup:** See [authentication](../_shared/authentication.md). For long jobs, use [async task polling](../_shared/async-tasks.md) with `POST /minimax/tasks`.
|
|
16
|
+
|
|
17
|
+
## Contract
|
|
18
|
+
|
|
19
|
+
| Parameter | Values | Default |
|
|
20
|
+
| --- | --- | --- |
|
|
21
|
+
| `model` | `minimax-h3` | `minimax-h3` |
|
|
22
|
+
| `prompt` | non-empty string | conditional |
|
|
23
|
+
| `image_urls` | 1–9 public HTTP(S) URLs | omitted |
|
|
24
|
+
| `audio_urls` | 1–3 public HTTP(S) URLs | omitted |
|
|
25
|
+
| `ratio` | `16:9`, `9:16` | `16:9` |
|
|
26
|
+
| `duration` | integer 4–15 | 4 |
|
|
27
|
+
| `async` | boolean | false |
|
|
28
|
+
| `callback_url` | public HTTP(S) webhook | omitted |
|
|
29
|
+
|
|
30
|
+
At least one of `prompt`, `image_urls`, or `audio_urls` is required. Mode inference is deterministic:
|
|
31
|
+
|
|
32
|
+
1. `audio_urls` present → audio-guided video
|
|
33
|
+
2. otherwise `image_urls` present → image-to-video
|
|
34
|
+
3. otherwise → text-to-video
|
|
35
|
+
|
|
36
|
+
Public price is **$0.25 per generated second**. Failed tasks are not charged.
|
|
37
|
+
|
|
38
|
+
## Text to video
|
|
39
|
+
|
|
40
|
+
```bash
|
|
41
|
+
curl -X POST https://api.acedata.cloud/minimax/videos \
|
|
42
|
+
-H "Authorization: Bearer $ACEDATACLOUD_API_TOKEN" \
|
|
43
|
+
-H "Content-Type: application/json" \
|
|
44
|
+
-d '{
|
|
45
|
+
"model": "minimax-h3",
|
|
46
|
+
"prompt": "A red fox running through a snowy forest at dawn, low tracking shot",
|
|
47
|
+
"ratio": "16:9",
|
|
48
|
+
"duration": 6,
|
|
49
|
+
"async": true
|
|
50
|
+
}'
|
|
51
|
+
```
|
|
52
|
+
|
|
53
|
+
## Multi-image video
|
|
54
|
+
|
|
55
|
+
```json
|
|
56
|
+
{
|
|
57
|
+
"model": "minimax-h3",
|
|
58
|
+
"prompt": "Preserve the character and clothing while the camera slowly pushes in",
|
|
59
|
+
"image_urls": [
|
|
60
|
+
"https://cdn.acedata.cloud/b1c82e4937.png",
|
|
61
|
+
"https://cdn.acedata.cloud/eb75d88a3f.png"
|
|
62
|
+
],
|
|
63
|
+
"ratio": "9:16",
|
|
64
|
+
"duration": 8,
|
|
65
|
+
"async": true
|
|
66
|
+
}
|
|
67
|
+
```
|
|
68
|
+
|
|
69
|
+
## Audio-guided video
|
|
70
|
+
|
|
71
|
+
```json
|
|
72
|
+
{
|
|
73
|
+
"model": "minimax-h3",
|
|
74
|
+
"prompt": "A dancer moves naturally to the rhythm",
|
|
75
|
+
"image_urls": ["https://cdn.acedata.cloud/b1c82e4937.png"],
|
|
76
|
+
"audio_urls": ["https://cdn.acedata.cloud/6f7d62b18b.wav"],
|
|
77
|
+
"ratio": "9:16",
|
|
78
|
+
"duration": 8,
|
|
79
|
+
"async": true
|
|
80
|
+
}
|
|
81
|
+
```
|
|
82
|
+
|
|
83
|
+
## Poll a task
|
|
84
|
+
|
|
85
|
+
```bash
|
|
86
|
+
curl -X POST https://api.acedata.cloud/minimax/tasks \
|
|
87
|
+
-H "Authorization: Bearer $ACEDATACLOUD_API_TOKEN" \
|
|
88
|
+
-H "Content-Type: application/json" \
|
|
89
|
+
-d '{"action":"retrieve","id":"TASK_ID"}'
|
|
90
|
+
```
|
|
91
|
+
|
|
92
|
+
Continue polling about every five seconds until `response.success` is true or an error appears. Use `retrieve_batch` with `ids` to check several tasks in one request.
|
|
93
|
+
|
|
94
|
+
## Gotchas
|
|
95
|
+
|
|
96
|
+
- Do not send `action` to `/minimax/videos`; the API infers the mode from media inputs.
|
|
97
|
+
- `duration` must be an integer, not a decimal.
|
|
98
|
+
- Audio mode takes precedence when both image and audio arrays are present.
|
|
99
|
+
- Use public URLs that the generation service can download.
|
|
100
|
+
- Returned videos are served from AceDataCloud CDN.
|
|
@@ -36,13 +36,13 @@ After every navigation, upload, tab switch, modal open/close, or submission, the
|
|
|
36
36
|
3. The page opens on `上传视频` by default. Select mode by exact visible tab text — `上传图文`, `上传视频`, or `写长文` — by observing the tab strip and clicking the returned ref. Verify the selected mode after clicking; the upload area text changes (`拖拽视频到此或点击上传` for video, an image dropzone for `上传图文`).
|
|
37
37
|
4. If the tab click reports success but the mode does not change, an onboarding popover is covering the tab strip. Press `Escape`, re-observe, and click again. If it still does not change, stop and ask the user to dismiss the overlay in the visible tab — never try to delete page nodes.
|
|
38
38
|
5. Upload one approved resource at a time and wait until the visible preview count reaches the submitted count before the next resource (up to 60 seconds per image). Re-observe between images: the file input's ref changes after the first upload. For video, wait until processing completes and Publish becomes enabled, up to 10 minutes. If resource resolution is unavailable, wait for the user to select local media and verify the same preview/processing state.
|
|
39
|
-
6. Fill the image/video title using the visible title textbox (recognition hints: placeholder containing `填写标题`, then the single visible title input fallback). Titles are capped at 20 full-width-equivalent characters; a visible `n/20` counter turning over the limit is authoritative, so shorten and re-confirm rather than submitting a truncated title. Fill body in the visible rich-text editor (`输入正文描述` placeholder). Use `browser.fill` for replacement, or `browser.click` followed by `browser.type` for rich text. Read immediately after each field
|
|
39
|
+
6. Fill the image/video title using the visible title textbox (recognition hints: placeholder containing `填写标题`, then the single visible title input fallback). Titles are capped at 20 full-width-equivalent characters; a visible `n/20` counter turning over the limit is authoritative, so shorten and re-confirm rather than submitting a truncated title. Fill body in the visible rich-text editor (`输入正文描述` placeholder). Use `browser.fill` for replacement, or `browser.click` followed by `browser.type` for rich text. Read immediately after each field and compare the complete title/body with the normalized preview. If a write fails or the content differs, do not append more keystrokes: replace the whole field, read it again, and obtain renewed confirmation for any changed preview. Never publish duplicated, missing, or extra text.
|
|
40
40
|
7. Limit tags to the first 10 confirmed tags. Insert them one at a time, close any topic suggestion popover by focusing the title, and verify visible chips/text before continuing.
|
|
41
41
|
8. Configure options one at a time and verify each exact state: schedule (1 hour–14 days), visibility (`公开可见`, `仅自己可见`, `仅互关好友可见`), originality, and products. If originality was requested but cannot be confirmed, abort rather than publishing non-original. Bind a product only when the exact intended product is visibly selected; never accept a first fuzzy match silently.
|
|
42
42
|
9. For long article: choose `写长文` → `新的创作`; fill `输入标题` textarea and the body editor; click `一键排版`; enumerate visible template names; select the confirmed template and verify its selected state; click `下一步`; then fill the separate publish-page description editor.
|
|
43
43
|
10. Before the final action, read or screenshot again and compare media count, title, full body, tags, options, products, and schedule with the confirmed preview. Stop on mismatch.
|
|
44
|
-
11.
|
|
45
|
-
12. Follow [reconciliation](./reconciliation.md). Immediate success requires
|
|
44
|
+
11. After the confirmed preview, run a fresh `browser.find` for `role=button`, `name=发布`. The current declarative creator widget is exposed through that semantic contract even though its page element is custom. Inspect the returned ref and reject `submit-disabled=true`, `disabled`, `aria-disabled=true`, or disabled styling. Click that fresh ref exactly once; do not substitute another visible control when Publish is missing.
|
|
45
|
+
12. Follow [reconciliation](./reconciliation.md). Immediate success requires `/publish/success`, `published=true`, or a newly added same-title entry/note ID in note management. A changed draft count, returning to the initial publish page, or click completion alone is not success. Return the canonical note URL when visible.
|
|
46
46
|
|
|
47
47
|
Never mix image/video media unless the visible current UI explicitly supports it. Bind products only when the account visibly exposes the feature and the exact selected products appear in the final preview.
|
|
48
48
|
|