maximize_misery

A goal-assessment skill distilled from CGP Grey's video "You're Probably Doing All 7" — a satirical guide to maximizing misery. Tactic 5 of the seven contrasts two ways of setting goals (SMART vs. VAPID); this skill turns that contrast into a practical goal-quality check.

The skill

assess-goal scores a goal on five axes, each running from a SMART pole (supports follow-through) to a VAPID pole (predicts drift), then classifies the goal and rewrites it sharper:

Specific     ↔  Vague
Measurable   ↔  Amorphous
Actionable   ↔  Pie-in-the-sky
Responsible  ↔  Irrelevant
Time-bound   ↔  Delayed

SMART and VAPID aren't two separate tests — they're the two ends of the same axes. A VAPID trait is just a broken SMART dimension. So one pass tells you both what's good and what's broken, and where the goal sits on the spectrum. (Originally these were two overlapping skills, assess-smart-goal and assess-vapid-goal; they were merged because they measured the same five axes.)

The recurring fix: shrink the goal to the smallest real step — "do this one pile of laundry," not "clean the whole house today."

A calibration corpus pins example goals to each score boundary so any future change to the rubric that shifts a classification gets caught.

Note: this is CGP Grey's wording for SMART (A = Actionable, R = Responsible), which differs from the classic Achievable/Relevant variant.

Installing the skill

Copy (or symlink) the skill folder into your skills directory:

cp -r skills/assess-goal ~/.claude/skills/
# or, to keep it live with this repo:
ln -sfn "$PWD/skills/assess-goal" ~/.claude/skills/assess-goal

Source material

The talk/ folder holds the full distillation of the video, produced with the talk-distill pipeline:

• captions.SRT — raw auto-captions (en-orig)
• transcript.txt — clean speaker-labelled transcript
• executive_summary.md — first-read overview of all 7 tactics
• key_takeaways.md — timestamp-indexed navigation bullets