tweet-rl-tracker_skill

This skill helps you build a Notion-based tweet performance tracker with a poor man's reinforcement loop, including automatic tweet screenshot capture.
  • TypeScript

186

GitHub Stars

1

Bundled Files

2 months ago

Catalog Refreshed

4 months ago

First Indexed

Readme & install

Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.

Installation

Preview and clipboard use veilstrat where the catalogue uses aiagentskills.

npx veilstrat add skill different-ai/agent-bank --skill tweet-rl-tracker

  • SKILL.md11.7 KB

Overview

This skill creates and manages a Notion-based tweet performance tracking system to implement a manual feedback loop for improving tweet quality over time. It includes automated tweet screenshot capture (including video frame capture at 10s) via Chrome DevTools MCP and a CLI-first workflow for adding entries to your Notion database. The system makes it easy to log inputs, measure outcomes, and run simple reinforcement-style experiments.

How this skill works

The tool sets up a Notion page and database with fields for tweet text, likes, impressions, calculated engagement score, hook type, topic, posted date, and optional media. A Chrome MCP workflow navigates to a tweet URL, extracts text/author/metrics, captures a screenshot (or a 10s video frame), uploads the image to an external host, and creates a Notion entry with the metadata and screenshot URL. Regular logging and weekly reviews reveal what hooks and topics perform best.

When to use it

  • When you want to learn which tweet hooks and topics drive engagement
  • To run simple, repeatable A/B-style experiments on tweet formats
  • If you need a lightweight RL-like feedback loop without complex tooling
  • When you want visual archives of tweets (screenshots or video frames)
  • For teams or creators tracking growth, virality, or conversation metrics

Best practices

  • Log tweets consistently after 24–48 hours to let metrics stabilize
  • Decide hook type before posting and record it as an input feature
  • Use the binary "Worked?" checkbox to simplify weekly analysis
  • Group and filter by Hook and Topic during review sessions
  • Host screenshots on a public file service (S3, Cloudinary) before adding to Notion

Example use cases

  • A founder testing whether questions or bold claims get more replies
  • A creator tracking which behind-the-scenes topics drive profile clicks
  • A marketer measuring amplification rate for campaign tweets
  • An analyst archiving tweet visuals and metadata for monthly hypothesis testing
  • An agency running weekly reviews to advise content strategy

FAQ

Yes. The Chrome profile used by MCP must be logged in to access some tweets and media reliably.

How do screenshots get into Notion?

Chrome MCP returns base64 images; save them to a public host (S3, Cloudinary, etc.), then add the hosted URL to the Notion files field.

Built by
VeilStrat
AI signals for GTM teams
© 2026 VeilStrat. All rights reserved.All systems operational