distill
Distill
You're over-specifying. The artifact tells the reader how to format output instead of what to accomplish. That scaffolding was necessary for weaker models. It constrains stronger ones.
The Problem
When writing skills, prompts, or instructions, you default to showing exactly what the output should look like — filled-in example tables, prescribed column names, template formats. This feels helpful but it creates a ceiling. The model matches your template instead of thinking about what the specific problem needs.
The Test
For each instruction in the artifact, ask: is this intent or format?
| Intent (keep) | Format (cut) |
|---|---|
| Summarize the results in a table | Use these columns: Name, Status, Duration, Owner |
| Assessments should be opinionated | Acceptable values: "ready", "blocked", "needs work" |
| Group items by category | Categories: Infrastructure, Frontend, Backend, Testing |
| Commit your changes after each step | Run git add -A && git commit -m "step N: description" |
More from corygabrielsen/skills
mission-control
Coordinate complex multi-step work using task graphs and parallel background agents. Use when work requires decomposition, delegation, and long-running operations that may survive context compaction.
11loop-codex-review
Automated code review loop with progressive reasoning levels. Runs n parallel Codex reviews (configurable via -n), Claude addresses issues, climbs from low→xhigh reasoning until fixed point (all n clean). Human approval at each iteration.
11review-pr
Thorough, educational PR review process. Gathers context, explains changes, identifies issues systematically, and drafts constructive feedback collaboratively before submitting.
11orthogonalize-pr
Identify and separate orthogonal (independent) change sets within a decomposed PR. Creates branches for each set that could theoretically be reviewed or merged separately.
10decontextualize
Course-correct when artifacts overfit to conversation context. Invoke when examples, names, or language are too specific to current work.
10loop-review-skill-until-fixed-point
Iterate /review-skill on a target until fixed point. Runs review passes until all reviewers return NO ISSUES.
9