The Agent Skills Directory

[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection because it processes untrusted content from external sources.
Ingestion points: Reads content from user-uploaded .zip files (SKILL.md Step 1) and fetches data from external conference URLs using WebFetch (SKILL.md Step 6).
Boundary markers: The instructions do not define boundary markers or provide explicit warnings to the agent to disregard instructions embedded within the processed LaTeX source files or website text.
Capability inventory: The skill has significant capabilities, including file extraction, searching, directory creation, and file manipulation using shell commands (unzip, grep, find, mkdir, cp, rm, cat) throughout SKILL.md.
Sanitization: There is no evidence of sanitization or filtering of the content extracted from the external files or URLs before it is processed by the agent.
[COMMAND_EXECUTION]: The skill relies on executing various shell commands to manage the file system and extract metadata from the LaTeX templates.
Evidence: Uses unzip to extract contents to /tmp, find and grep to identify LaTeX components and document classes, and mkdir, cp, and rm for restructuring the output directory and cleaning up temporary files.

latex-conference-template-organizer