The Agent Skills Directory

[DYNAMIC_EXECUTION]: The skill constructs and executes shell commands by interpolating user-provided input into a command string.
Evidence: In SKILL.md, the commands rcs ask claude "{{ARGUMENTS}}" and claude -p "{{ARGUMENTS}}" directly place the {{ARGUMENTS}} placeholder within a shell command.
Risk: This pattern is vulnerable to command injection if the user input contains shell metacharacters such as backticks, semicolons, or dollar signs, potentially allowing for arbitrary command execution on the host system.
[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted user data which is then used to parameterize high-capability shell commands.
Ingestion points: The {{ARGUMENTS}} variable in SKILL.md captures input from the user prompt.
Boundary markers: Absent. There are no delimiters or instructions to ignore embedded commands within the user input.
Capability inventory: The skill utilizes shell command execution via the claude and rcs CLI tools.
Sanitization: Absent. The skill wraps the input in double quotes but does not perform escaping or validation of the content before it is processed by the shell.

ask-claude