awesome-openclaw-skills

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

• Third-party content exposure detected (high risk: 0.85). The required runtime workflow is installing/using third-party OpenClaw skills from the ClawHub/skills registry (e.g., via clawhub install author/skill), which causes the agent to ingest that outsider-authored skill’s readable files (like skill.yaml, README.md, and code/docs) into its LLM context as part of executing the skill.

MEDIUM W012: Unverifiable external dependency detected (runtime URL that controls agent).

• Potentially malicious external URL detected (high risk: 0.80). The README explicitly instructs that pasting a GitHub skill URL into the OpenClaw chat (e.g. https://github.com/openclaw/skills/tree/main/skills/steipete/slack) causes the assistant to fetch and install that repository at runtime, which means remote skill code and skill.yaml from that URL can be executed and thereby control the agent's prompts/behavior.