Skip to content
Back to skills

107 Instructions Example Fc7c9897

ASecurity

Demonstrates using instructions (system prompts) to control LLM behavior

  • 9 stars
  • 0 votes
  • 0 copies
  • 0 views
  • Added October 11, 2026
ai-agents

Security analysis

A100/100

Scanned October 11, 2026

npx -y skills add tools-only/X-Skills --skill 107-instructions-example_fc7c9897 --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of 107 Instructions Example Fc7c9897?

Add the live security badge to your README. It updates with every re-scan.

Security grade badge for 107 Instructions Example Fc7c9897
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/tools-only-107-instructions-example-fc7c9897/badge)](https://www.skillsdirectory.com/skills/tools-only-107-instructions-example-fc7c9897)

More formats (shields.io, HTML) on the badges page. Keep it an A: scan every change in CI with Pro.

SKILL.md
---
name: Instructions Example
description: Demonstrates using instructions (system prompts) to control LLM behavior
---

# Instructions Example

This document demonstrates how to use the `instructions` parameter to control LLM behavior with system prompts.

## Source Content

{{ ref('source.md').content }}

---

## Without Instructions

**Summary**: {{ ref('source.md') | llm_extract('a brief one-sentence summary of the feedback') }}

---

## With Pirate Instructions

**Pirate Summary**: {{ ref('source.md') | llm_extract('a brief one-sentence summary of the feedback', instructions='Talk like a pirate. Use "Arrr" and pirate vocabulary.') }}

---

## LLM Block with Instructions

{% llm instructions='Talk like a pirate. Use "Arrr" and pirate vocabulary.' %}
Summarize this product feedback in one sentence: {{ ref('source.md').content }}
{% endllm %}

---

## Using Classification from Another Model

{% llm instructions='You are a customer success manager.' %}
Based on this sentiment analysis:

{{ ref('classify.md').content }}

Write a one-sentence professional response acknowledging the feedback.
{% endllm %}

Attribution

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments

Loading comments…

Newsletter

Get the weekly list of safe new skills, and the ones we pulled.