security-first-methodology

What problem does it solve?

This Skill addresses the need for a security-first approach in autonomous agent workflows, ensuring that security is integrated into all stages of development.

Core Features & Use Cases

• Threat Modeling: Adds threat modeling to plans, security gates before execution, and adversarial review for security-critical decisions.
• Recipe-aware Constraints: Applies constraints based on the task type, such as threat modeling for new projects and pre-execution gate checks.
• Security Patterns: Provides guidelines for code generation, such as avoiding hardcoded secrets and validating inputs at trust boundaries.
• Second-opinion Review: Offers an optional review process for high-risk tasks, including cross-review with Claude Code or Codex, manual fresh-chat review, and local adversarial review.

Quick Start

Use the security-first-methodology skill to activate the adaptive security-first methodology for your autonomous agent workflow.