The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The utility script find-polluter.sh is susceptible to argument injection via the TEST_PATTERN variable. A maliciously crafted pattern could be used to execute unintended find primary actions.
Ingestion points: User-supplied test pattern argument in find-polluter.sh.
Boundary markers: None present in the script code.
Capability inventory: Execution of find and npm test subprocesses in find-polluter.sh.
Sanitization: Lacks sanitization for shell metacharacters or dangerous find arguments (e.g., -delete) in the pattern input.
[COMMAND_EXECUTION]: The skill includes a bash script that automates the execution of local test files using npm test. While this is part of the intended debugging functionality, it represents a capability that iterates over and executes project files.
[SAFE]: The debugging instructions and documentation encourage secure practices, such as verifying the existence of environment variables and secrets without exposing their literal values in logs.

systematic-debugging