The Agent Skills Directory

[REMOTE_CODE_EXECUTION]: The skill provides commands for the agent to install the xparse-cli tool by piping remote scripts directly into the shell using source <(curl ...) and irm ... | iex. This allows for the execution of arbitrary code from a remote server without verification.- [EXTERNAL_DOWNLOADS]: The installation section references external scripts hosted on dllf.intsig.net. Downloading and executing scripts via shell pipes is a high-risk activity that can be exploited for remote code execution.- [COMMAND_EXECUTION]: The skill instructs the agent to execute the xparse-cli binary and manage temporary files. While it includes safety practices like restricted file permissions (0600) and avoiding echo for sensitive data, the core functionality relies on potentially dangerous shell interactions.- [PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection because it reads and processes content from external PDF and image files.
Ingestion points: read_content, read_pages, and search_text commands extract text from user-provided files.
Boundary markers: There are no instructions or delimiters provided to help the agent distinguish between the document's content and the agent's actual tasks.
Capability inventory: The agent can execute shell commands via xparse-cli and manage files, creating a path for malicious instructions to trigger actions.
Sanitization: The skill does not describe any methods for filtering or sanitizing document content before it is processed.

xparse-doc-tools