The Agent Skills Directory

COMMAND_EXECUTION (SAFE): find-polluter.sh executes npm test on local files to identify which test creates side effects. This is a standard development task.
COMMAND_EXECUTION (SAFE): Methodology examples demonstrate using git, codesign, and security commands for diagnosing build and environment issues within the local workspace.
DATA_EXFILTRATION (SAFE): The skill includes a defensive programming example for verifying the existence of environment secrets without logging their literal values (using ${IDENTITY:+SET}), reducing risk of accidental exposure.
INDIRECT_PROMPT_INJECTION (LOW): As a debugging tool, the skill processes external data such as error logs and test outputs. While this represents a theoretical injection surface, the skill focuses on analytical reasoning and manual verification, which mitigates the risk of automated obedience to embedded instructions.

systematic-debugging