Skip to content
Get packs

Score to a Target

Scores work out of 10 with a fixed rubric, then improves it until every item hits your target score.

Source Download skill
Repository
MengTo/Skills
Stars
7,187
Licence
MIT
Files
3 files

What it does

When you ask an agent to rate something out of 10 and reach a bar such as 8, this skill makes it write a clear scoring rubric first, break each item into small scored criteria, and score from real evidence like screenshots or measurements rather than guesses. It then improves the weakest items round by round, re-scores them with the same rubric, and reports honestly which items still fall short. It also suggests independent or blind judging and saving a scorecard for future work.

Great for

  • Game or app developers who want every asset or screen to reach a set quality bar
  • Designers and artists who compare their work against named benchmarks and need consistent scores across rounds
  • Teams that want a documented before-and-after scorecard of quality improvements

Skip it if

  • Quick subjective opinions where a rigorous multi-round process is overkill (the workflow is heavy by design)
  • Work that cannot be captured as renders, recordings or measurements, since the skill says scores should come from real evidence and not from code or memory
  • Situations where a score is used to make high-stakes decisions about people, since the scores are admitted judgements

Try asking

  • “Score these five character animations out of 10 and get every one to at least 8”
  • “Rate each screen in this app out of 10 and bring anything under 6 up to 6 or more”
  • “What would make this landing page a 9 out of 10? Score it against the benchmark sites I named”
Step-by-step guide: how to use Score to a TargetInstall, first run, examples, tips and FAQs

Install in 3 steps

  1. 1

    Download the skill

    The ZIP contains the 3 original files, the MIT licence and an attribution note.

    workflow-score-to-target.zip
  2. 2

    Add it to your agent

    Claude Code (all projects)

    mkdir -p ~/.claude/skills && unzip workflow-score-to-target.zip -d ~/.claude/skills/

    Claude Code / Agent SDK (this project only)

    mkdir -p .claude/skills && unzip workflow-score-to-target.zip -d .claude/skills/

    Claude apps (web/desktop): Settings → Capabilities → Skills → Upload skill, and choose the ZIP.

  3. 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

  • Name the benchmark you judge against (for example specific games or sites) so scores have a reference point.
  • Give the target and say whether it applies to every item or the average, since the skill treats 'get them to 8' as every item reaching 8.
  • Pair it with a screenshot or capture workflow so scores come from actual renders rather than the agent's memory of the work.

What's inside

  • SKILL.md
  • templates/critic-prompt.md
  • templates/scorecard.md

Redistributed unmodified from MengTo/Skills 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.