The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to execute a Python script located at a path that resolves outside the skill's own directory through the use of directory traversal pointers. The files checks, commands, modules, scripts, and support all contain relative path references (../../) that map the skill's components to the parent directory environment. This allows the skill to execute code and access modules from the broader filesystem beyond its own root.
[PROMPT_INJECTION]: The skill exposes an indirect prompt injection surface by ingesting documentation from the filesystem without security controls.
Ingestion points: PRD documentation files located in the docs/prd-helper directory (specified by the --docs-root argument in the status command).
Boundary markers: Absent. The skill does not use delimiters or instructions to ignore embedded commands within the ingested PRD data.
Capability inventory: The skill has access to the Bash tool, allowing for shell command execution.
Sanitization: None. There is no evidence of validation or sanitization of the documentation content before it is processed by the agent.

prd-status