The Agent Skills Directory

[COMMAND_EXECUTION]: The skill performs shell script execution and git operations using parameters provided by the user.
Evidence: Step 3 executes a local script ~/.claude/skills/init-latex-project/scripts/init.sh and Step 4 invokes init-python-project, both of which process user-supplied project names and paths within a shell environment.
Evidence: The skill uses git remote add and git push commands with SSH URLs provided by the user in Step 1 to configure repository remotes.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface by incorporating unsanitized user descriptions into project files.
Ingestion points: The 'Research summary' field in Step 1 collects free-form text from the user.
Boundary markers: Absent. The summary is directly embedded into the PROJECT.md file and used to pre-fill content in LaTeX repository files without delimiters.
Capability inventory: The skill possesses Bash, Write, and Edit permissions, allowing it to execute system commands and modify the file system based on context derived from the summary.
Sanitization: Absent. No validation, escaping, or filtering is performed on the user-provided research summary before it is written to the file system.

project-init