full-output-enforcement

Enforces complete code generation by banning placeholder patterns and truncation.

Updated May 11, 2026
One-click install
npx skills add https://github.com/cloudofgeorge/AI-hands-Engineer --skill full-output-enforcement-cloudofgeorge
Or copy as Structured Prompt for Agentβ–Ό
Please help me install this Agent Skill.
Skill: full-output-enforcement
Source: https://github.com/cloudofgeorge/AI-hands-Engineer/tree/main/skills/design/full-output-enforcement
Command: npx skills add https://github.com/cloudofgeorge/AI-hands-Engineer --skill full-output-enforcement-cloudofgeorge

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? LLMs often truncate long outputs, insert placeholder comments like "// rest of code", or ask "want me to continue?" instead of delivering complete work. This Skill overrides that default behavior so every requested deliverable is generated in full. ## Core Features & Use Cases - Placeholder Ban: Hard-fails on patterns like // ..., // TODO, "for brevity", and skeleton-only outputs. - Scope Locking: Counts expected deliverables before generating, then cross-checks the output against that count. - Token-Limit Handling: Pauses at clean breakpoints with a resumable [PAUSED β€” X of Y complete] marker instead of compressing or skipping content. - Use Case: Ask for a full 5-component React dashboard; the Skill ensures all 5 components arrive as complete, runnable code with no omitted sections. ## Quick Start Apply full-output enforcement and generate the complete implementation of all requested files with no placeholders or omissions.

Frequently Asked Questions about full-output-enforcement

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

FAQPage Schema
How do I stop an LLM from truncating code output?β–Ό

Apply explicit anti-truncation rules that ban placeholder patterns like "// rest of code" and "for brevity". This Skill locks the deliverable count upfront and cross-checks the output so nothing is skipped or shortened.

How to handle token limits during long code generation?β–Ό

Write at full quality up to a clean breakpoint such as the end of a function or file, then emit a pause marker like "[PAUSED β€” X of Y complete]". On "continue", resume exactly where you stopped without recap or repetition.

What output patterns indicate incomplete LLM responses?β–Ό

Common signs include comments like "// ...", "// implement here", "// similar to above", and prose like "let me know if you want me to continue" or "the rest follows the same pattern". Structural shortcuts like skeleton-only output are also failures.

When should I not use full-output enforcement?β–Ό

Avoid it for exploratory brainstorming, quick prototypes, or when you explicitly want a short summary or skeleton. The Skill treats every task as production-critical, which adds length and latency unnecessary for casual queries.

Does this work with any LLM or coding assistant?β–Ό

Yes, it is a prompt-level instruction set with no code dependencies, so it works with any model that accepts system or skill instructions. Effectiveness depends on the model's instruction-following ability and context window size.