The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a shell command (mkdir -p ./audit-reports/{target-slug}) using a slug derived from user-supplied $ARGUMENTS. While the instructions suggest converting the argument to a URL-safe name, this relies on the agent's interpretation and creates a risk of command injection if the input contains shell metacharacters that are not effectively sanitized before execution.
[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection as its primary function is to ingest and analyze untrusted text provided by users.
Ingestion points: Written content, copy, and UI text provided via $ARGUMENTS for audit and review (SKILL.md).
Boundary markers: Absent; there are no instructions or delimiters used to ensure the agent treats the content strictly as data and ignores any embedded instructions.
Capability inventory: The skill has the capability to create directories (mkdir) and write files to the local file system (saving reports in audit-reports/ and planning-docs/).
Sanitization: Absent; the skill does not specify any validation or filtering logic for the input text before processing.

david-copy