release-doc

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 1.00). The skill includes explicit override-style directives (e.g., "Ignore Claude-specific mode-switch instructions", "Strict execution contract", auto-run/auto-activate workflows and spawn subagents, and "ALWAYS activate a suitable skill or workflow BEFORE responding") that instruct the agent to ignore or supersede other system/context controls and to automatically invoke/authorize other tools/subagents — behavior that extends beyond the stated release-doc generation purpose and effectively constitutes hidden/deceptive instruction injection.