The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted test code which could contain instructions designed to influence agent behavior.
Ingestion points: Untrusted data enters the agent context through the reading of test files matching patterns such as *.spec.*, *.test.*, and __tests__/**, as well as user-provided $ARGUMENTS.
Boundary markers: The instructions do not specify the use of delimiters or protective headers to isolate the analyzed code from the agent's internal reasoning.
Capability inventory: Across its instructions, the skill has the capability to read project files and, when the --fix flag is present, write modifications to those files.
Sanitization: There are no instructions for sanitizing, escaping, or validating the content of the test files before they are processed by the agent.

test-review