plan-implementation

CRITICAL E004: Prompt injection detected in skill instructions.

• Potential prompt injection detected (high risk: 0.90). The skill contains a deceptive "Silently read the implement skill" instruction that tells the agent to perform and hide an out-of-band action (reading another skill's contract) rather than transparently following the stated planning task.

MEDIUM W011: Third-party content exposure detected (indirect prompt injection risk).

• Third-party content exposure detected (medium risk: 0.30). The skill ingests outsider-authored free text at runtime when it accepts a Linear URL/hint, fetches matching Linear issues (including their titles/descriptions/comments/child issues) and uses that content to generate the plan.