The Agent Skills Directory

[COMMAND_EXECUTION]: The skill constructs a shell command by directly interpolating user input ($ARGUMENTS) into a command string: node ${CLAUDE_PLUGIN_ROOT}/src/digest.js $ARGUMENTS. This creates a vulnerability to command injection if the user provides input containing shell metacharacters (e.g., semicolons, pipes, or ampersands), allowing the execution of arbitrary shell commands.
[INDIRECT_PROMPT_INJECTION]: The skill lacks explicit sanitization or delimitation logic when handling external data. * Ingestion points: User input from $ARGUMENTS and output from the digest.js script (SKILL.md). * Boundary markers: Delimiters and instructions to ignore embedded instructions are absent. * Capability inventory: Bash tool used for executing Node.js scripts (SKILL.md). * Sanitization: No evidence of validation, escaping, or filtering of the processed content exists.

cco-digest