> 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 search recent Lemon8 videos and archive one traceable result page?

> Query recent Lemon8 videos with lemon8.app.fetch_search (main, video, recent), validate and archive raw data with archive_search.py, and link each shortlisted row to its archive file and request ID.

To search recent Lemon8 videos, call `lemon8.app.fetch_search` with `search_tab` `main`, `filter_type` `video` and `order_by` `recent`. Save the response, check and archive the response with `archive_search.py`, and record the request parameters in a collection manifest. Then hand over a shortlist in which every row links to the archive file and the request ID.

For example, you are planning a post about reapplying sunscreen, but “skincare” is too broad to turn into a shoot. Search one page of recent videos and look for situations you can demonstrate, such as reapplying at work, outdoors or over makeup. Lemon8's [creator guide](https://www.lemon8-app.com/experience/features-and-uses-of-lemon8-app-for-creators?region=us) also uses search to find formats worth studying; use the same approach to find questions your own post can answer.

## How do I turn my search results into a brief for next week's shoot?

1. Search recent videos for `sunscreen reapplication`.
2. Give your AI assistant a sheet with the shortlisted links and what you observed in each video.
3. Ask: “Group these videos by reapplication scenario. Suggest one question I can demonstrate for each group. Tell me what the source actually shows, what I still need to verify, and which link backs up each idea.”

A video that only shows the finished makeup works as a visual reference. A video that shows the timing and the steps helps you design your own comparison. If a result only has a title, mark the result for review and decide whether to fetch the post detail.

## 1. Choose a keyword and budget

Start with one page for one keyword. Every changed keyword, every additional page and every post-detail request is billed separately. [BeatAPI pricing](https://beatapi.io/pricing)

## 2. Retrieve a filtered page

```bash
curl --fail-with-body -sS https://api.beatapi.io/v1/social-data/call \
  -H "Authorization: Bearer $BEATAPI_API_KEY" \
  -H 'Content-Type: application/json' \
  -d '{"action":"lemon8.app.fetch_search","params":{"query":"sunscreen reapplication","search_tab":"main","filter_type":"video","order_by":"recent","max_cursor":""}}' \
  -o lemon8-search-page-1.json
```

Replace the `query` value with your keyword and keep the English values `main`, `video` and `recent`. If the HTTP request fails, keep the error and stop. An HTTP success still needs a business-status check in step 3.

## 3. Validate and archive

Save the script below as `archive_search.py`. BeatAPI sanitizes and passes through the search `data` directly, so `archive_search.py` preserves the complete `data` payload and `request_id` without assuming a uniform list path. `archive_search.py` refuses to overwrite an existing archive file.

```python
import json, pathlib, sys
source, target = map(pathlib.Path, sys.argv[1:3])
r = json.loads(source.read_text(encoding="utf-8"))
if r.get("status") != "succeeded" or r.get("action") != "lemon8.app.fetch_search":
    raise ValueError("Expected successful Lemon8 search")
if "data" not in r or r["data"] is None:
    raise ValueError("Missing query data")
record = {"action": r["action"], "request_id": r.get("request_id"), "data": r["data"]}
with target.open("x", encoding="utf-8") as f:
    json.dump(record, f, ensure_ascii=False, indent=2)
print(str(target))
```

## 4. Inspect the output

```bash
python3 archive_search.py lemon8-search-page-1.json lemon8-search-archive-1.json
```

Check that the archive has the target action and the preserved raw `data`. In a collection manifest, record `query`, `filter_type`, `order_by`, `search_tab`, the collection time and the requested `max_cursor`. A successful empty result is still a search outcome; an empty result does not prove that no matching content exists on Lemon8.

## 5. Continue deliberately and inspect selected posts

1. Read the returned `max_cursor` and `has_more`, and save both values.
2. Keep the filters unchanged and save each page under a new filename.
3. Stop when the cursor is repeated or missing, when `has_more` is false, or when you reach the page budget.

A post-detail query for a selected post is an extra call. Search results are not a comment collection; comments also require `group_id`, `item_id` and `media_id`.

## 6. Hand off the shortlist and its sources

Hand over the original response together with your research sheet. For example, for a carry-on packing review, use rows like the rows below to record why each item stays and what to check next. The short labels are row references only; keep the actual post ID and the original link in the same row.

| Ref | What you saw                                     | Decision and next step                                              |
| --- | ------------------------------------------------ | ------------------------------------------------------------------- |
| L01 | A demo of packing winter clothes into a carry-on | Keep the video and check the case dimensions and number of garments |
| L02 | A shot of an already-packed case                 | Keep as a visual reference and look elsewhere for the steps         |
| L03 | Only the title was available                     | Save the title and source, then look up the post if selected        |

Link each row to the archive file and the request ID, and spot-check that one row leads back to the original post. When the search comes back empty, save the filters and the response anyway.

[Filter parameters](/faq/lemon8-search-filters) · [Links and comments](/faq/lemon8-links-comments)