mirror of
https://github.com/BagelHole/DevOps-Security-Agent-Skills.git
synced 2026-08-22 12:49:53 +02:00
33 lines
1000 B
Markdown
33 lines
1000 B
Markdown
---
|
|
name: llm-app-security
|
|
description: Secure LLM-powered applications with input validation, output controls, tenant isolation, and abuse prevention.
|
|
license: MIT
|
|
metadata:
|
|
author: devops-skills
|
|
version: "1.0"
|
|
---
|
|
|
|
# LLM Application Security
|
|
|
|
Harden chatbots and AI features embedded in web and mobile products.
|
|
|
|
## Baseline Security Checklist
|
|
|
|
- Validate and classify all user-provided context.
|
|
- Separate system prompts from user content strictly.
|
|
- Add moderation for toxic, harmful, and policy-violating outputs.
|
|
- Enforce tenant boundaries in retrieval and memory layers.
|
|
- Rate-limit high-cost endpoints.
|
|
|
|
## Secure RAG Pattern
|
|
|
|
1. Ingest content with malware and secret scanning.
|
|
2. Tag documents by tenant and access policy.
|
|
3. Filter retrieval candidates by user authorization.
|
|
4. Add provenance metadata in final responses.
|
|
|
|
## Related Skills
|
|
|
|
- [ai-agent-security](../ai-agent-security/) - Agent-specific controls
|
|
- [sast-scanning](../scanning/sast-scanning/) - Secure coding checks
|