behavioral-interview-prepper

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

• Third-party content exposure detected (high risk: 0.85). The skill’s runtime workflow explicitly reads hojicha/candidate-context.md (Step 1) and then “probe[s] for real experiences” and “append[s] all new stories to hojicha/candidate-context.md” (Step 2.5), where the candidate’s free-form responses are outsider-authored text that can be ingested into the agent’s LLM context for later answer generation.