The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests and processes user-supplied input via the $ARGUMENTS variable and passes it directly to internal workflows.
Ingestion points: The $ARGUMENTS variable in SKILL.md captures all user input provided to the skill.
Boundary markers: There are no explicit delimiters or boundary markers used to separate the routing logic from the content of the arguments.
Capability inventory: The skill is configured to use powerful tools including Bash, Write, and Read as specified in the allowed-tools metadata.
Sanitization: No sanitization, validation, or escaping of the user-provided arguments is performed before they are passed to the new-workspace or remove-workspace workflows, which may lead to unintended command execution or instruction overrides if the sub-workflows do not handle the data securely.

gsd-workspace