> For clean Markdown of any page, append .md to the page URL.
> For a complete documentation index, see https://docs.beatapi.io/llms.txt.
> For AI client integration (Claude Code, Cursor, etc.), connect to the MCP server at https://docs.beatapi.io/_mcp/server.

# How do I retrieve Weibo video details and check play-count scope for a weekly report?

> Retrieve verified video-post details, establish the actual metric and original/repost relationship, and preserve comparable observations and missing fields.

Pass the verified video-post ID as `mid` to `weibo.app.fetch_video_detail` to retrieve a Weibo video's details. For a weekly play-count report, check each returned field and the video the field belongs to, then compare only the same metric across videos. Record reposts and original videos separately.

For example, for three product demos, record the verified play counts, interactions and collection times side by side. If one video returns only interactions, leave the video's play count blank, then review the video and its comments for improvements.

## How do you retrieve a known video post?

Pass the verified video-post identifier as `mid` to `weibo.app.fetch_video_detail`:

```json
{
  "action": "weibo.app.fetch_video_detail",
  "params": {"mid": "<confirmed_video_post_id>"}
}
```

Send the JSON to `POST https://api.beatapi.io/v1/social-data/call` with normal BeatAPI Bearer authentication. `weibo.app.fetch_video_detail` takes `mid`, not a user uid and not pasted share text. See [post identifiers](/faq/weibo-identifiers).

Reuse saved details first. To pick candidate videos from an account, call `weibo.web_v2.fetch_user_posts` with `feature: 3`, which selects video-related extended data, and read the returned posts. `feature: 3` does not guarantee a complete list of original videos and does not export play counts.

## What needs checking before the report?

| Check                         | Recording approach                                                                   |
| ----------------------------- | ------------------------------------------------------------------------------------ |
| Which object owns the count   | Preserve confirmed video-post, original-video and repost relationships               |
| Plays, reading or interaction | Retain the actual metric name and meaning rather than merging every views-like value |
| Original value and unit       | Save the source value and establish display units before conversion                  |
| Observation time              | Record retrieval time and differing publication ages                                 |
| Missing or failed reads       | Separate missing, unread and failed states; retain a reported zero as zero           |

Do not assume another service's `metrics.views` as a fixed field here: the local `result_schema` of `weibo.app.fetch_video_detail` has no play-count path. Until you review a real response, label the report's play-count column “awaiting field verification,” not zero. Likes, comments and reposts do not substitute for a missing play count.

## How do comparable counts inform content decisions?

1. **Compare like with like.** Compare the same kind of object under the same metric definition, with clear observation times.
2. **Separate video ages.** Compare this week's new videos separately from videos that have accumulated plays for days.
3. **Record changes with both values.** For a change, keep both original values, both times and the sources, and describe the result as the net change observed over that interval.
4. **Read the strong demos.** For higher-play demonstrations, read the video and the discussion: which question the opening answers, whether the steps are clear and what commenters still ask. Put that evidence in the content suggestions.

A play count alone does not show completed viewing or purchases.

Prompt for the assistant:

> Confirm the original video posts in this list and preserve IDs, publication times and sources. Reuse details, verify actual play-count fields and units, and separate missing or unverified values before comparing. Read the demonstrations and comment questions to explain approaches worth studying next.

## What can the handoff contain now?

Deliver the verified video list, the original detail responses, the metric names and units, the observation times and the suggestions that the content supports. With the handoff, the team can choose demonstrations worth studying and see which material is still missing.

Budget the selected video-detail requests with current [pricing](https://beatapi.io/pricing). Comments are additional research material; reuse existing results for analysis.