The Problem

Content creators and developers need a centralized, searchable list of Claude Opus 5.5 prompts that produced viral videos. Without a curated source, teams waste time hunting for working prompts and tracking which video belongs to which prompt.

What This Does

The repository stores 703 Markdown files in prompts/, each containing a prompt and metadata for a single video. data/videos.json aggregates the metadata (video IDs, categories, links). README.md presents a curated gallery of the 418 videos that have live remakes on Skillry, with links to the original posts and prompt files. The collection is constantly updated by contributors.

How It Is Wired

There is no executable code. The project is a static content store.

  • Entry point – README.md (or ./prompts/*.md) serves as the main interface.
  • Data flow – No runtime data manipulation; all assets are read‑only.
  • External touchpoints – The repository references external URLs (Skillry, Twitter) but does not perform network calls or write to a database.
  • File responsibilities
  • README.md – human‑readable index and gallery.
  • prompts/*.md – individual prompt documents.
  • data/videos.json – machine‑readable index for tooling.
  • LICENSE, .gitignore – standard repository packaging.

Because there is no code graph, there are no call chains, no shared state, and no side effects beyond the file system.

How To Use It

  1. Clone ``bash git clone https://github.com/moses-y/awesome-opus5-5-videos cd awesome-opus5-5-videos ``
  2. Explore – Open README.md for a visual gallery or prompts/*.md for raw prompt text.
  3. Integrate – Copy the Markdown content into your own documentation or feed it into a prompt‑management tool that can parse Markdown.
  4. Consume – Use data/videos.json to programmatically search by category or ID. ``python import json, pathlib videos = json.loads(pathlib.Path('data/videos.json').read_text()) ``

No build or runtime configuration is required.

Real‑World Use

A marketing team can import data/videos.json into a CMS, automatically generating a gallery page that lists each video with its prompt and original author. Developers can import a prompt file and paste it into their own Claude Opus 5.5 session to recreate or remix the animation.

Code Health & Issues

  • Missing CI/CD – No .github/, .gitlab-ci.yml, or other pipeline config.
  • Sparse testing – Only one test file exists, unrelated to the prompts collection.
  • No automation for updates – Adding a new prompt requires manual edits to Markdown and JSON.
  • No licensing for prompts – Individual prompt ownership is unclear; contributors may need to verify rights before redistribution.

The Bottom Line

The repo is a well‑structured, static catalog of Claude Opus 5.5 prompts and associated videos. It lacks any executable logic, CI pipelines, or automated update mechanisms. Teams that need a ready‑to‑use prompt library should find it useful, but they will need to add their own tooling if they want to automate imports or enforce quality checks.