security-hardening

SkillSecurity

This skill adds a security layer to your AI agent. With it in place, your agent blocks prompt injection attempts, validates the inputs it receives, and sanitizes its outputs. It can also run tasks in a sandbox and use STRIDE threat modeling to check for weaknesses.

Available today. Use it from your connected AI after setup.

Add the skill to your agent, then keep using it as you normally would. The security checks run in the background with no extra setup.

Then ask your AI: use the security-hardening skill

What your AI can do with it

  • Block prompt injection attempts before they affect your agent
  • Validate incoming inputs for unsafe content
  • Run agent tasks inside a sandbox so risky operations stay contained
  • Sanitize outputs before they leave your agent
  • Apply STRIDE threat modeling to spot potential weaknesses

What this skill tells your AI

The instructions your AI receives, as published by a5c-ai/babysitter in library/methodologies/ruflo/skills/security-hardening/SKILL.md and read by ahel’s review.

  • When processing untrusted inputs
  • Security audits of agent-generated code
  • Compliance verification (OWASP Top 10, CIS)

AIDefence Layers

  1. Prompt Injection Detection - Pattern + heuristic blocking
  2. Input Validation - Path traversal, type coercion, parameter sanitization
  3. Static Analysis (SAST) - Vulnerability scanning, CWE matching
  4. Sandboxed Execution - Network isolation, filesystem restrictions, resource limits
  5. Output Sanitization - Secrets, PII, injection vector redaction

Security Levels

LevelLayersUse Case
standardSAST + validation + sanitizationRoutine audits
elevated+ threat modeling + compliancePre-release audits
maximum+ sandbox + full STRIDE + remediationCritical systems

Agents Used

  • agents/security-auditor/ - Vulnerability detection
  • agents/reviewer/ - Code quality verification

Tool Use

Invoke via babysitter process: methodologies/ruflo/ruflo-security-audit

Signals

GitHub stars
2k
Forks
106
Last commit
Sep 2026
Advanced
Catalog kind
skill
Gateway key
security-hardening
Source
github.com/a5c-ai/babysitter