title-research

Generate YouTube titles by mining what actually performs, not by guessing.

Pipeline: research → extract patterns → generate → score → report.

Depends on:

• firecrawl CLI (authenticated)
• the title-score skill, installed as a sibling directory of this one

Paths below are relative to this skill's base directory, not the user's project. Resolve them from the SKILL.md you are reading, e.g.:

SKILL_DIR=<absolute path of the directory containing this SKILL.md>
SCORE="$SKILL_DIR/../title-score/scripts/score.py"