Blind Prediction Logger
Writes an immutable, blind-scored prediction log for a final video script before publishing.
- Repository
- XBuilderLAB/cheat-on-content
- Stars
- 7,443
- Licence
- MIT
- Files
- 1 file
What it does
This skill guides an AI agent through recording a prediction of how a finished video script will perform, before it is published, so later results can be compared against it. The agent reads the script and a scoring rubric, has a separate sub-agent score it without outside context, then drafts a probability distribution and scenarios for the user to review. Once approved, the prediction is saved as a file that is meant to be locked against edits, and it refuses to make predictions if the user has already seen post-publication data.
Great for
- Creators who want an honest calibration loop comparing predicted and actual video performance
- People who already use the cheat-on-content workflow and its scripts/ and rubric_notes.md files
Skip it if
- Anyone wanting a guaranteed accurate forecast: the output is an AI judgement and is only as good as the rubric and calibration data
- Users who have already seen views or likes: the skill is designed to refuse or downgrade such predictions, so results would be reconstructed rather than blind
Try asking
- “start prediction for scripts/2025-03-01_a1_intro.md”
- “给这稿子打分并预测”
- “写预测日志 for my final cut before I post it”
- “Run a v2 prediction after I rewrote the script post-shoot”
Install in 3 steps
- 1
Download the skill
The ZIP contains the 1 original files, the MIT licence and an attribution note.
cheat-predict.zip - 2
Add it to your agent
Claude Code (all projects)
mkdir -p ~/.claude/skills && unzip cheat-predict.zip -d ~/.claude/skills/Claude Code / Agent SDK (this project only)
mkdir -p .claude/skills && unzip cheat-predict.zip -d .claude/skills/Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.
- 3
Ask for the task
No special command needed — the agent reads the skill's description and loads it when your request matches. Try one of the prompts above.
Get better results
- Only run it on the final script you will actually publish, and answer the 'have you seen any data' question honestly.
- Review the drafted scores and bucket distribution rather than just typing ok, since the tool is designed around your corrections.
- Keep rubric_notes.md and calibration samples up to date so the confidence level is meaningful.
What's inside
- SKILL.md
Redistributed unmodified from XBuilderLAB/cheat-on-content under the MIT licence. The assessment was drafted with AI from the skill's files on 2026-10-29. Skills can instruct your agent to run code — read SKILL.md and any scripts before installing.