skills.rest
Debindenny
DebindennyCOMMUNITY
Explore all skills by this author →
GitHub

full-output-enforcement

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

Updated Aug 11, 2026
Search Tags:#prompt engineering#code generation#llm output#anti-truncation#placeholder ban#complete output
One-click install
npx skills add https://github.com/Debindenny/travelplanner-community- --skill full-output-enforcement-debindenny
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: full-output-enforcement
Source: https://github.com/Debindenny/travelplanner-community-/tree/main/travel-planner/agent/skills/full-output-enforcement
Command: npx skills add https://github.com/Debindenny/travelplanner-community- --skill full-output-enforcement-debindenny

SYSTEM DOCUMENTATION & REQUIREMENTS

What problem does it solve? LLMs frequently 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. - Scope Locking and Cross-Check: Counts expected deliverables before generating, then verifies the count before responding so nothing is omitted. - Token-Limit Handling: Pauses at clean breakpoints with a resumable marker instead of compressing or skipping content when output approaches the token limit. - Use Case: When asking an AI to generate five complete React components or a full configuration file, apply this Skill to receive all five finished implementations rather than one example plus descriptions of the rest. ## 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 fully. The model counts expected deliverables first, then cross-checks completeness before responding.

How to get complete code instead of placeholders from AI?▼

Instruct the model to treat partial output as failure and prohibit patterns such as "// TODO", "// similar to above", and "for brevity". Require runnable code in every block rather than descriptions of what the code would do.

What happens when the output hits the token limit?▼

The model writes at full quality up to a clean breakpoint such as the end of a function or file, then ends with a pause marker stating progress. Sending "continue" resumes exactly from that point with no recap or repetition.

Which placeholder patterns are banned in generated code?▼

Banned patterns include "// ...", "// rest of code", "// implement here", "// TODO", "/* ... */", bare ellipses, and prose like "the rest follows the same pattern" or "and so on" replacing actual content.

When should I not use full-output enforcement?▼

Avoid it for quick exploratory questions, summaries, or brainstorming where brevity is the goal. It is designed for tasks requiring exhaustive deliverables such as full files, complete component sets, or unabridged documentation.

Related Skills

Explore More Software Engineering Skills→
darkmatterdarkmatter
1

definition-of-done

Define explicit completion criteria for ambiguous agentic tasks.

Official
Intermediate
PeterBookerPeterBooker
23

test

Execute Go unit tests for specified or all packages.

Community
Basic
tiankong0101-bytetiankong0101-byte

verify

Verify task outputs for correctness, completeness, and quality assurance.

Community
Basic
richardwhiteiirichardwhiteii
50

RLM (Recursive Language Model) Skill

Load large text outside the prompt, chunk it, and run batched sub-queries with aggregation.

Community
Advanced