The Agent Skills Directory

[COMMAND_EXECUTION]: The utility script scripts/run-tests.sh uses the eval command to execute an xcodebuild command string that is dynamically constructed from user-supplied arguments.
Evidence: In scripts/run-tests.sh, the variable CMD is built by concatenating parameters such as SCHEME, DEVICE, and TEST_CLASS without shell escaping, and then executed via eval "${CMD}".
Risk: This pattern is vulnerable to command injection. An attacker could provide arguments containing shell metacharacters (e.g., ;, &&, or backticks) to execute arbitrary commands on the host system.
[PROMPT_INJECTION]: The skill presents an indirect prompt injection surface where untrusted data could influence agent actions through the test-runner script.
Ingestion points: Script arguments in scripts/run-tests.sh including --scheme, --device, --class, and --method.
Boundary markers: The script does not utilize boundary markers or explicit instructions to prevent the interpretation of embedded shell commands in the arguments.
Capability inventory: The skill includes the capability to execute subprocesses and system commands via eval within scripts/run-tests.sh.
Sanitization: There is no evidence of input validation, sanitization, or shell-safe escaping of parameters before they are interpolated into the execution string.

swift-tdd-workflow