create-evidence-case

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow explicitly calls the /dogpile Tier‑3 research tool to run web searches (see SKILL.md "If INCONCLUSIVE → Tier 3 Research" which invokes .claude/skills/dogpile/run.sh), meaning the agent ingests open/public search results (untrusted third‑party content) and re-evaluates/verdicts based on that content.