skills.rest
ConnectiveTCS
ConnectiveTCSCOMMUNITY
Explore all skills by this author →
GitHub

prompt-engineering

Automate secure prompt design and safe task routing for LLM assistants.

Updated Apr 5, 2026
Search Tags:#security#prompt engineering#injection prevention#guardrails#task routing#output validation#llm safety
One-click install
npx skills add https://github.com/ConnectiveTCS/Gradient_Generator --skill prompt-engineering-connectivetcs
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: prompt-engineering
Source: https://github.com/ConnectiveTCS/Gradient_Generator/tree/main/.agents/skills/prompt-engineering
Command: npx skills add https://github.com/ConnectiveTCS/Gradient_Generator --skill prompt-engineering-connectivetcs

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes references (resource) components.

What problem does it solve?

Prompt engineering and task routing for secure, reliable LLM-assisted workflows, focusing on guardrails, injection prevention, and validated outputs.

Core Features & Use Cases

  • Secure system prompt design and clear separation between system and user content
  • Injection detection, prompt sanitization, multi-step orchestration, and output validation
  • Task routing, intent classification, and safe tool calling with robust safety checks
  • Use Case: For JARVIS-like assistants, route complex tasks with safeguards and sandboxed tool calls.

Quick Start

Set up a secure system prompt for a given task and run a sanitized user input through the routing workflow.

Frequently Asked Questions about prompt-engineering

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

FAQPage Schema
How do I prevent prompt injection in LLM-driven assistants?▼

Task routing for multi-step LLM workflows uses intent classification and tool call allowlists to direct actions safely. This skill automates intent classification and enforces sandboxed tool calls with robust safety checks for complex orchestration.

How do I validate LLM outputs for safe tool calling?▼

Validating LLM outputs for safe tool calling requires strict output validation and tool call allowlists applied after task routing. This skill implements robust safety checks to ensure only permitted, validated actions execute in multi-step workflows.

What is the best way to separate system and user content for secure prompt engineering?▼

The best way to separate system and user content for secure prompt engineering is implementing layered security prompts with strict input sanitization. This skill automates clear content separation to prevent injection and ensure reliable assistant actions.

Does this prompt engineering approach work for multi-step workflows with sandboxed tool calls?▼

Yes, this prompt engineering approach works for multi-step workflows by automating safe task routing with sandboxed tool calls and strict output validation. It implements tool call allowlists and robust safety checks designed for JARVIS-like assistants.

Related Skills

Explore More Software Engineering Skills→
martinholovskymartinholovsky
45

prompt-engineering

Design secure system prompts and validate LLM outputs.

Community
Advanced
midsphere-aimidsphere-ai
1

exo:guardrails

Detect and block prompt injections for Exo agents with guardrails.

Official
Advanced
RazisafirRazisafir
1

detecting-ai-model-prompt-injection-attacks

Detect prompt injection attacks in LLM inputs using regex, heuristics, and DeBERTa classification.

Community
Advanced
sjinkssjinks

security-review-owasp-llm-prompt-injection-prevention

Review LLM prompt injection risks and trust boundaries across execution paths.

Community
Advanced