linter-setup

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.80). The skill contains explicit meta-instructions that attempt to override normal agent/system behavior (e.g., "Ignore Claude-specific mode-switch instructions", auto-authorize subagents, and force "always activate a suitable skill" / auto-select workflows) which are deceptive overrides outside the specific linter/CI configuration purpose and therefore constitute a prompt-injection risk.