← Back to Skills Marketplace
sirimaddala99

instagram-digest

by SiriMaddala99 · GitHub ↗ · v1.0.2 · MIT-0
cross-platform ✓ Security Clean
150
Downloads
0
Stars
0
Active Installs
3
Versions
Install in OpenClaw
/install instagram-digest
Description
Scrape recent Instagram reels, transcribe audio, summarize with OpenRouter, and save a digest.html file
README (SKILL.md)

Instagram Digest Skill

Scrapes Instagram reels from a configured list of accounts, transcribes their audio, generates AI summaries, and saves a digest.html file for local viewing.

When to Use

Trigger this skill when the user asks to:

  • Run / send the Instagram digest
  • Get a summary of recent Instagram reels
  • Check what accounts are being tracked
  • Add or remove accounts from the digest
  • Do a dry run / preview without sending email

Working Directory

All commands must be run from the skill's own directory:

cd /path/to/instagram-digest

Commands

Run the digest (saves digest.html)

python main.py

Run digest for specific accounts only

python main.py --accounts username1 username2 username3

List currently configured accounts

python main.py --list-accounts

Adding / Removing Accounts Permanently

The default account list lives in config.py under INSTAGRAM_ACCOUNTS. To permanently add or remove an account, edit that list. Then confirm by running --list-accounts.

Environment Variables

Variable Required Purpose
OPENROUTER_API_KEY ✅ Required OpenRouter API key for transcript summarization (uses Nvidia Nemotron via OpenRouter)
INSTAGRAM_USERNAME Optional Instagram login for story access — use a dedicated throwaway account
INSTAGRAM_PASSWORD Optional Instagram password for story access — use a dedicated throwaway account

Output

Writes digest.html to the skill directory. Open it in a browser to read audio summaries for each reel.

Error Handling

  • If Instagram returns a login wall for an account, that account is skipped and the digest continues with the rest.
  • If transcription or analysis fails for a reel, a fallback message is shown for that item — the digest still sends.
  • If no new reels or stories are found across all accounts, the email is skipped entirely and a message is printed.
Usage Guidance
This skill appears to do what it says, but review and consider the following before installing or running it: - You must provide an OpenRouter API key; transcripts and captions are sent to OpenRouter for summarization. - If you supply Instagram credentials, the skill will save session cookies to .instagram_session.json and reuse them — consider using a dedicated throwaway Instagram account as the SKILL.md suggests. - The code downloads media (yt-dlp) and may download Whisper model weights at first run (faster-whisper), which can consume disk space, bandwidth, and time; run in a controlled environment (virtualenv, container) if concerned. - There are minor packaging/documentation inconsistencies (duplicate/unused entries in requirements files, SKILL.md marked instruction-only despite included scripts) — review the source if you want full assurance. - If you don't trust the source, run the code in an isolated sandbox, inspect the code yourself, or avoid providing any real Instagram credentials. If you proceed, consider creating a Python virtual environment and installing dependencies manually so you control what gets installed.
Capability Analysis
Type: OpenClaw Skill Name: instagram-digest Version: 1.0.2 The instagram-digest skill is a legitimate tool for scraping Instagram reels, transcribing their audio, and generating AI-powered summaries. It utilizes Playwright for browser automation, yt-dlp for media retrieval, and faster-whisper for local transcription, with summaries generated via the OpenRouter API. The handling of Instagram credentials and session persistence (scripts/scraper.py) is consistent with the stated purpose of bypassing login walls, and no evidence of data exfiltration, malicious execution, or prompt injection was found.
Capability Assessment
Purpose & Capability
The name/description match the code and runtime requirements: scraping (playwright, requests), downloading media (yt-dlp), local transcription (faster-whisper), and summarization via OpenRouter (OPENROUTER_API_KEY). The optional INSTAGRAM_USERNAME/INSTAGRAM_PASSWORD env vars are reasonable for avoiding login walls. Minor note: top-level requirements.txt includes 'anthropic' which is unused by the code; scripts/requirements.txt lists 'openai' (used) — this packaging inconsistency is likely benign but sloppy.
Instruction Scope
SKILL.md instructs running the bundled Python scripts from the skill directory (cd + python main.py) which matches the included code. The code will: (1) launch Playwright to scrape Instagram pages (may open a visible browser on login), (2) save session cookies to .instagram_session.json if you log in, (3) download thumbnails and reels via HTTP/yt-dlp, (4) transcribe using a Whisper model and (5) send transcripts to OpenRouter for summarization. These actions are within the stated purpose but result in local persistence of session cookies and outbound network traffic to Instagram, package/model hosts, and OpenRouter — the SKILL.md mostly documents this but you should expect these side effects.
Install Mechanism
There is no automated install spec; the project includes requirements files listing packages to pip install. Runtime behavior will also pull potentially large model weights via faster-whisper/WhisperModel('base') and yt-dlp will fetch media from web hosts — both can download significant data at runtime. No arbitrary archive downloads or suspicious external installers are present, but expect model and media downloads.
Credentials
The single required credential (OPENROUTER_API_KEY) is necessary for the summarization step and is declared as primary. INSTAGRAM_USERNAME/INSTAGRAM_PASSWORD are optional and used only to obtain a session for story/login-wall access; no other unrelated credentials or secrets are requested.
Persistence & Privilege
The skill writes files in its directory and temporary directories: digest.html, .instagram_session.json (persisted session cookies when you log in), and temporary media/model files (tmp directories cleaned after run, but the session file persists). always is false and the skill does not request elevated platform privileges. The persistent session file can contain authentication state — the README suggests using a throwaway account.
How to Use
  1. Make sure OpenClaw is installed (local or Docker)
  2. Run the install command in chat: /install instagram-digest
  3. After installation, invoke the skill by name or use /instagram-digest
  4. Provide required inputs per the skill's parameter spec and get structured output
Version History
v1.0.2
- Added scripts/requirements.txt to specify Python dependencies. - Added scripts/scraper.py for Instagram scraping functionality. - Added scripts/transcriber.py for audio transcription tasks.
v1.0.1
- Updated from Claude API to OpenRouter (Nvidia Nemotron) for transcript summarization, requiring OPENROUTER_API_KEY. - Environment variable requirements and documentation updated accordingly. - Requirements file moved from scripts/requirements.txt to root requirements.txt. - Improved environment variable descriptions, marking Instagram credentials as optional. - Minor description and metadata adjustments for clarity.
v1.0.0
Initial release of instagram-digest: - Scrapes recent Instagram reels from a configurable list of accounts. - Transcribes audio from each reel. - Summarizes audio content using Claude and saves results in a local digest.html file. - Supports listing, adding, and removing tracked accounts through config.py. - Handles login walls and per-reel errors gracefully, always producing a digest when possible. - Can preview, customize accounts for each run, and optionally send results by email.
Metadata
Slug instagram-digest
Version 1.0.2
License MIT-0
All-time Installs 0
Active Installs 0
Total Versions 3
Frequently Asked Questions

What is instagram-digest?

Scrape recent Instagram reels, transcribe audio, summarize with OpenRouter, and save a digest.html file. It is an AI Agent Skill for Claude Code / OpenClaw, with 150 downloads so far.

How do I install instagram-digest?

Run "/install instagram-digest" in the OpenClaw or Claude Code chat to install it in one step — no extra setup required.

Is instagram-digest free?

Yes, instagram-digest is completely free, licensed under MIT-0. You can download, install and use it at no cost.

Which platforms does instagram-digest support?

instagram-digest is cross-platform and runs anywhere OpenClaw / Claude Code is available (cross-platform).

Who created instagram-digest?

It is built and maintained by SiriMaddala99 (@sirimaddala99); the current version is v1.0.2.

💬 Comments