The Agent Skills Directory

[COMMAND_EXECUTION]: The Update Mode workflow generates and executes Bash commands (such as git log and grep) using variables like <related_files> and <feature>. These variables are parsed directly from the contents of markdown request documents, which can be influenced by any user able to edit files in the repository.
[PROMPT_INJECTION]: The skill is vulnerable to indirect prompt injection (Category 8) because it ingests untrusted data from the workspace to perform sensitive operations.
Ingestion points: Reads existing request documents from the docs/features/{feature}/requests/ directory.
Boundary markers: There are no explicit delimiters or instructions provided to the agent to ignore or escape instructions contained within the analyzed files.
Capability inventory: The skill is granted Bash, Write, Read, Grep, and Glob tools, providing a significant footprint for potential exploitation.
Sanitization: The instructions do not define any sanitization, escaping, or validation logic for the content extracted from files before it is passed to the Bash tool for execution.

create-request