instagram-reel-index
Instagram Reel Index
A Claude Code skill that turns any Instagram creator’s reels into one searchable overview. For every reel you get:
- the title and what it covers
- what you get out of it, and the tools it names
- the comment keyword for the free resource, with one-click copy
- views, likes, comments and length
Search by tool (“MCP”, “Higgsfield”), sort by views, open the reel.
Ready-made example: @nateherkai, 114 reels (4 Jul to 4 Oct 2026)
▶ Open it in your browser: https://buildmeight-beep.github.io/instagram-reel-index/examples/nateherkai/overview-standalone.html
| File | What |
|---|---|
examples/nateherkai/overview-standalone.html |
The same overview as one file to keep (5.2 MB, covers inside). Open the link, click Download raw file, then double-click the file |
examples/nateherkai/overview.xlsx |
Excel with cover pictures, filters on every column, and a Tools sheet ranking every tool by how many reels name it |
Unofficial index of public posts, not made or endorsed by Nate Herk.
Make your own for any account
What you need
| What | Cost | |
|---|---|---|
| 1 | Claude Code (any paid Claude plan) | Your normal plan. Claude writes the titles and descriptions, which uses your usage limit |
| 2 | Python 3 | Free. For the Excel file also run once: pip install openpyxl pillow |
| 3 | Node.js (only to install the Glasser command) | Free, nodejs.org |
| 4 | A Glasser account with credit: glasser.ai | Pay per call. The only paid outside tool. It fetches the posts and reel transcripts from Instagram |
You do not need Apify, an Instagram login, or any other API key.
My experience with Glasser: I’m testing it. You get $1 of free credit when you sign up, which already covers about 500 reels. I added another $10 to keep going.
What it cost me
The @nateherkai example above, 114 reels:
| Item | Amount |
|---|---|
| Glasser calls | 125 (10 pages of posts + 114 transcripts + 1 test) at $0.00188 each |
| Glasser total | $0.235 |
| Per reel | about $0.002 (0.2 cents); 100 reels ≈ $0.20 |
| Claude | inside my normal plan, no extra bill. The two biggest writing steps (titles from the cover images, “what you get”) took about 270,000 Sonnet tokens |
The skill tells you the expected cost and waits for your OK before it spends anything. Every call is logged with its price in charges.log.
Install
- On this page click Code → Download ZIP and unzip it.
- Copy the folder
instagram-reel-indexinto your Claude Code skills folder:- Windows:
C:\Users\<you>\.claude\skills\ - Mac/Linux:
~/.claude/skills/
- Windows:
- Install the Glasser command once:
npm install -g @glasser-ai/cli@latest, then runglasser balanceand log in when it asks. - Start Claude Code in any folder and say: “Make an Instagram reel index for @nateherkai for the last 3 months”
You get overview.html, overview-standalone.html, overview.xlsx and overview.csv in instagram-index/<handle>/.
Good to know
- Public posts only. Use it for your own research.
- Covers are downloaded on the day you scrape, because Instagram’s image links expire after a few days. The script does this for you.
- Re-running continues where it stopped and does not pay twice for what is already downloaded.
How it works
| Step | Script | What it does |
|---|---|---|
| 1 | scripts/scrape.py |
Posts, reel transcripts and covers through Glasser (ScrapeCreators); logs every charge |
| 2 | scripts/prepare.py |
Bundles caption, transcript and the caption keyword per reel |
| 3 | Claude | Writes notes.json: title (read from the cover), what it covers, what you get, tools, keyword, freebie |
| 4 | scripts/build.py |
Builds the HTML overview and CSV (--embed for the single-file version) |
| 5 | scripts/excel.py |
Builds the Excel file |
Full instructions for Claude are in instagram-reel-index/SKILL.md.