The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute shell commands, specifically ln -sfn and ls -la, to manage symbolic links within the local filesystem.
[COMMAND_EXECUTION]: The skill modifies the contents of ~/.claude/skills/, which is a sensitive directory containing the definitions of the agent's capabilities and behaviors. Arbitrary modification of this directory can alter the agent's operational logic.
[PROMPT_INJECTION]: The skill uses user-supplied skill names to construct filesystem paths for command execution. If the input is not strictly validated, it could potentially be exploited for path traversal attacks (e.g., using ../../ in a skill name) to link or overwrite unintended directories.

skill-symlinker