full-output-enforcement

Enforces complete, unabridged LLM output by banning placeholder patterns and truncation.

Updated Apr 9, 2026
One-click install
npx skills add https://github.com/ivaylo91/Shopping-Mobile-App --skill full-output-enforcement-ivaylo91
Or copy as Structured Prompt for Agent
Please help me install this Agent Skill.
Skill: full-output-enforcement
Source: https://github.com/ivaylo91/Shopping-Mobile-App/tree/main/.agents/skills/full-output-enforcement
Command: npx skills add https://github.com/ivaylo91/Shopping-Mobile-App --skill full-output-enforcement-ivaylo91

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? LLMs often truncate long responses, insert placeholder comments like "// rest of code", or skip deliverables to save tokens, leaving users with incomplete code and partial answers. 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 "the rest follows the same pattern" in both code and prose. - Deliverable Cross-Check: Counts expected deliverables from the request, then verifies the output count matches before responding. - Clean Token-Limit Handling: When output approaches the token limit, it pauses at a clean breakpoint with a resumable marker instead of compressing or skipping content. - Use Case: Ask for a full multi-file project scaffold or five complete React components, and receive every file and component fully implemented with no skeletons or omissions. ## Quick Start Apply the full-output-enforcement skill 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?

Apply explicit output-enforcement rules that ban placeholder patterns like "// rest of code" and require every requested deliverable to be generated in full. This Skill also cross-checks the deliverable count against the original request before responding.

How to handle token limits during long code generation?

Write at full quality up to a clean breakpoint such as the end of a function or file, then emit a pause marker stating progress and the next section. On "continue", resume exactly where you stopped without recap or repetition.

What placeholder patterns should be banned in AI-generated code?

Ban patterns like "// ...", "// TODO", "// implement here", "// similar to above", and prose such as "for brevity" or "the rest follows the same pattern". These indicate omitted content and produce non-runnable output.

Does output enforcement work for non-code writing tasks?

Yes, the same rules apply to prose deliverables such as reports, documentation, and multi-section answers. The deliverable counting and banned-phrase checks ensure every requested section is complete.

When should I not use full-output enforcement?

Avoid it for quick exploratory questions, brainstorming, or cases where a short summary is explicitly wanted. Enforcing exhaustive output in those scenarios wastes tokens and slows down iteration.