build-openclaw-skill

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

• Third-party content exposure detected (high risk: 0.80). The skill's workflow and references explicitly involve installing and indexing skills from public sources—e.g., references/clawhub-publishing.md describes publishing/installing skills from GitHub/ClawHub (the agent/CLI parses SKILL.md from public repos) and references/metadata-gating.md documents installer entries with external download URLs—meaning the agent/runtime will fetch and parse untrusted third-party content that can influence tool actions.