> 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.

# Review Channels video comments

> Read a Channels video's top-level comments and reply threads with wechat_channels.v2.fetch_video_comments, then deliver a list of viewer questions the team can answer.

To find out what viewers still ask about a Channels video, request the video's comments with `wechat_channels.v2.fetch_video_comments`, using the numeric video ID as `object_id`. Start with the top-level comments. To open one reply thread, keep the same `object_id` and add the parent comment's `comment_id`. For an exhibition guide, for example, the deliverable is a list of questions the team should answer.

## Get the first comment page

If you already have the numeric video ID, use the numeric video ID directly. If you have a share link or an `export/…` identifier, first follow the [identifier workflow](/faq/wechat-channels-identifiers), and check that the returned title and account match the video you mean. Send the request below to `POST https://api.beatapi.io/v1/social-data/call` with BeatAPI Bearer authentication, and replace the example video ID with your own.

```json
{"action":"wechat_channels.v2.fetch_video_comments","params":{"object_id":"14941130915890399732","comment_id":"","last_buffer":"","raw":false}}
```

The top-level comments are in `data.comments`. `data.down_continue` tells you whether the same video and branch has another page; for the next page, pass back `data.last_buffer`. Start with one page, and retrieve at most two more pages when the questions need more context. Treat the comment count shown on the video as background only. In the report, state the pages actually retrieved, the number of usable comments and the reason you stopped.

## Restore discussion context

A comment such as “Do we enter here?” might refer to the side door shown in the tour, so the reply thread matters. Save the original wording and the video, then take the parent `comment_id` from the actual top-level response. Keep the same `object_id`, set `comment_id` to the parent ID and start with an empty `last_buffer`. Page through the reply thread with the thread's own responses, and save a separate manifest for the thread. Keep video IDs and parent comment IDs as numeric strings, so spreadsheet software preserves every digit.

```json
{"action":"wechat_channels.v2.fetch_video_comments","params":{"object_id":"14941130915890399732","comment_id":"<parent_comment_id>","last_buffer":"","raw":false}}
```

Export the saved pages with the [comment exporter](/faq/wechat-channels-comments), up to three saved pages per video and branch. Read the original wording, and the time and engagement values where available, in `comment-observations.csv`, then confirm the covered scope in `coverage.json`. When the same comment appears in several pages, keep the source of each observation, and count comments by confirmed comment identifiers.

## Deliver questions the team can answer

Give each row of the question list these columns: the question, the original wording, the video or reply location, the conditions, the answer status and the next step. For example, if a viewer asks about wheelchair access and the author's reply mentions the east ramp, record the author's reply and check the current venue information before you update the tour. When a viewer describes an earlier rule, record the time the rule applied.

Each first page, additional page and reply branch is a separate request. See current [pricing](https://beatapi.io/pricing).