
from correctless63
Goal-directed red team assessment skill for live systems with source-code access; enforces isolation and intensity gating before active testing.
Provides a structured, objective-driven red team workflow for live adversarial assessments where the operator has source code and live system access. Emphasises isolation checks, an intensity gate, boundary mapping, seam identification, attack execution, and mandatory progress visibility and reporting. Includes agent prompts, reporting format, and post-assessment actions.
Use this skill only for controlled, isolated penetration tests with explicit rules of engagement and test/sandbox credentials. Do NOT run against public internet targets or production systems. Activation requires a high-intensity authorization and environmental checks (private IPs, isolated VPS, or Docker Compose networks).
Intended for advanced coding/ops agents capable of orchestration and running background probes (Claude Code, Codex, or similarly capable automation agents). Human oversight strongly recommended for actual execution.
This skill has not been reviewed by our automated audit pipeline yet.