The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill instructs the user to install a CLI tool using a high-risk piped execution pattern: curl -fsSL https://cli.inference.sh | sh. This allows for the execution of unverified remote code within the user's shell environment.
[EXTERNAL_DOWNLOADS]: The skill downloads binary files from dist.inference.sh and uses npx to dynamically add additional skill dependencies from inference-sh/skills.
[COMMAND_EXECUTION]: The skill requests permission to execute arbitrary infsh commands via Bash(infsh *), which provides the capability to run a wide range of external applications and system operations.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection through its content sourcing functionality.
Ingestion points: External data is fetched from the web using tavily/search-assistant and exa/search tools referenced in SKILL.md.
Boundary markers: No delimiters or protective instructions are used to prevent the agent from following directives found in retrieved search results.
Capability inventory: The skill has permissions to execute CLI tools, write to social media via x/post-create, and generate images.
Sanitization: No sanitization or validation of the retrieved web content is performed before it is used to generate newsletter commentary or social media posts.

newsletter-curation