skills.rest
edivad1999
edivad1999COMMUNITY
Explore all skills by this author →
GitHub

principle-encode-lessons-in-structure

Converts recurring corrections into lint rules, metadata flags, runtime checks, or scripts.

1|Updated Aug 26, 2026
Search Tags:#code quality#automation#best practices#feedback loop#agent instructions#Ci enforcement#lint rules
One-click install
npx skills add https://github.com/edivad1999/stuc-stack --skill principle-encode-lessons-in-structure-edivad1999
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: principle-encode-lessons-in-structure
Source: https://github.com/edivad1999/stuc-stack/tree/main/skills/principle-encode-lessons-in-structure
Command: npx skills add https://github.com/edivad1999/stuc-stack --skill principle-encode-lessons-in-structure-edivad1999

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? Repeatedly writing the same textual instruction wastes effort and fails because readers must notice, remember, and comply. This Skill replaces recurring instructions with structural mechanisms that enforce the rule automatically. ## Core Features & Use Cases - Structural Encoding: Turns repeated corrections into lint rules, metadata flags, runtime checks, or automation scripts instead of more documentation. - Strongest-Rung Selection: Chooses the strongest available mechanism, from unrepresentable states that fail compilation down to runtime checks. - Feedback Loop: Captures every correction, routes it to the right layer (brain note, skill, lint rule, or principle), and closes the loop with concrete action. - Use Case: When you catch yourself telling an agent a second time not to use a banned API, encode it as a CI lint rule and delete the instruction. ## Quick Start Apply this principle whenever you notice a recurring correction by asking whether it can become a lint rule, metadata flag, runtime check, or script, then encode it and remove the textual instruction.

Frequently Asked Questions about principle-encode-lessons-in-structure

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I stop repeating the same instruction to an AI agent?▼

Encode the rule as a structural mechanism instead of text: a lint rule, metadata flag, runtime check, or script. Structural enforcement works without requiring the reader to notice, remember, and comply with written instructions.

What is the best way to enforce coding rules automatically?▼

Pick the strongest mechanism the situation allows: an unrepresentable state that cannot compile, then a lint or banned API that fails CI, then a canonical helper, then a runtime check. Agents copy surrounding code, so weaker guards become templates.

When should a correction become a lint rule versus a note?▼

Route by recurrence: one-off corrections go to a brain note, recurring fixes become skills or lint rules, and systemic issues become principles. If you write the same instruction a second time, it is a pattern worth encoding.

Why do textual coding instructions fail in practice?▼

Textual instructions require the reader to notice, remember, and comply, which makes them easy to miss. Structural mechanisms like lint rules and runtime checks enforce the rule without cooperation from the reader.

What are common mistakes when recording lessons from corrections?▼

Three anti-patterns: acknowledging without recording, recording without routing to implementation, and fixing one instance without generalizing the recurring pattern. Each leaves the underlying problem intact.

Related Skills

Explore More Software Engineering Skills→
itsadijmbtitsadijmbt
1

self-improvement

Capture agent learnings, corrections, and error context into structured logs.

Community
Intermediate
cursorcursor
6.6k

no-comments

Removes unjustified code comments by auditing a diff with a review subagent and fixing root causes.

Official
Advanced
imjasonhimjasonh
2

principle-encode-lessons-in-structure

Convert recurring corrections into lint rules, metadata flags, runtime checks, or scripts.

Community
Basic
yanochka11yanochka11

write-judge-prompt

Design LLM-as-Judge prompts with criteria, examples, and binary verdicts.

Community
Advanced