The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill fetches external content from a GitHub repository to provide up-to-date resources, creating a surface for indirect prompt injection.\n- Ingestion points: The skill instructs the agent to fetch the content of a README file from https://raw.githubusercontent.com/gmh5225/awesome-ai-security/refs/heads/main/README.md in SKILL.md.\n- Boundary markers: The instructions lack explicit delimiters or "ignore embedded instructions" warnings for the fetched data.\n- Capability inventory: The agent is expected to process the fetched markdown to identify security resources and links.\n- Sanitization: No sanitization or validation of the external content is described in the skill.\n- [EXTERNAL_DOWNLOADS]: The skill fetches supplemental resource links from the author's public repository on GitHub.\n- Fetches the README.md file from gmh5225/awesome-ai-security. This is a functional feature intended to keep the agent updated with current security tools and research.

llm-attacks-security