You're reading this. The skill predicted it. A workflow that turns every post into a calibrated experiment—score, blind-predict, retro, evolve. The future doesn't reward effort, it rewards those who see the pattern first. 1M followers in a month — not luck, system.
Cheat on Content is a workflow for content creators that records predictions and results for each post, reviews performance later, and updates the criteria used for future decisions. It is intended to make content planning and publishing an experiment that becomes more informed over time. The catalogue contains the skills that implement this workflow.
Skill Claude Code
A method for turning content ideas and published results into measurable predictions. It scores content, predicts its performance, reviews the result after three days, and updates the scoring rules; the built-in rules are for opinion videos.
Skill Claude Code
A workflow for changing a content-scoring rubric, meaning the rules and weights used to judge content. It supports either changing the rubric itself or recalculating only the performance buckets.
Skill Claude Code
An onboarding workflow for setting up a content-prediction project from scratch. It asks about the project, creates its folders and templates, and checks whether its automation hooks work.
Skill Claude Code
A workflow for learning patterns from another content creator’s scripts and performance data. It turns those examples into notes and evaluation rules for the cheat-on-content tool.
Skill Claude Code
A migration tool for updating an old .cheat-state.json file to the schema version expected by the current cheat-on-content tool. A schema is the format and set of fields that a data file must follow.
Skill Claude Code
A workflow that builds a description of who follows and responds to a content account. It derives the profile from comments and performance notes in earlier prediction records.
Skill Claude Code
A workflow that records an unchangeable prediction about how a piece of content may perform before later results are known. Blind prediction means judging the draft without looking at those later results.
Skill Claude Code
A small publishing-record procedure for prediction files, where it saves a published URL, platform, and time in metadata. It also updates the project's state file without changing the prediction text.
Skill Claude Code
A topic-selection workflow that reads a candidate list and recommends the highest-ranked video ideas. It uses existing scores, publication history, rejected ideas, safety filters, and recent topics to avoid duplicates.
Skill Claude Code
A review step that collects a prediction's later results, compares them with the original forecast, and records lessons for future scoring rules. T+N means checking performance N days after the prediction.
Skill Claude Code
INTERNAL sub-agent for blind 7-dim rubric scoring. NOT a user-facing skill — do NOT invoke from main conversation. Called via Task tool by cheat-score / cheat-predict / cheat-bump to get a context-isolated score on a script. Receives ONLY scriptpath + rubricnotespath; refuses any other input. Outputs strict JSON: 9…
Skill Claude Code
A rubric-based scoring step for one writing draft. A rubric is a set of named criteria used to judge work consistently.
Skill Claude Code
A topic-development conversation that starts from the user's own idea or experience, then narrows it to one angle and writes a draft. It can also run an optional batch mode for several topics.
Skill Claude Code
A workflow step for recording that a planned video has been filmed. It creates the video’s folder, saves the script used, checks whether the filmed version differs from the draft, and updates the publishing queue.
Skill Claude Code
A project progress dashboard that reads the current state and scans relevant files to show what work remains. It is designed for a prediction-and-review project.
Skill Claude Code
A tool that collects recent popular topics from configured sources such as Hacker News, Reddit, YouTube, and Bilibili, then prepares a shortlist of possible ideas.
At most 3 mods per repository are shown here, and a mod shipped inside a plugin is left to that plugin's page — the rest are on their repository pages: