The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill reads external guidelines and findings files to perform its tasks, creating a potential attack surface where malicious instructions could be embedded in those files.
Ingestion points: Reads guidelines and findings files as specified in the dispatcher prompt (Step 1).
Boundary markers: The instructions do not specify any delimiters or safety markers for the content of these files.
Capability inventory: The skill has the ability to modify source code, execute tests (via verify_commands), and create git commits.
Sanitization: There is no instruction to sanitize or validate the content of the findings or guidelines before processing.
[COMMAND_EXECUTION]: The skill executes shell commands for running tests and committing code. This behavior is expected and necessary for its intended purpose of automated refactoring.
Evidence: Use of verify_commands for test execution and git -c commit.gpgsign=false commit for version control in Step 3.

develop-refactor