The Agent Skills Directory

[COMMAND_EXECUTION]: The skill provides a shell script find-polluter.sh that identifies problematic test files by running npm test on files matched via a search pattern. Additionally, SKILL.md and root-cause-tracing.md include examples of executing system utilities like env, security, codesign, and git to collect diagnostic data and verify signatures.
[INDIRECT_PROMPT_INJECTION]: The skill incorporates "Pressure Test" scenarios in files such as test-pressure-1.md, test-pressure-2.md, and test-pressure-3.md. These files present adversarial environments (e.g., financial emergencies or social pressure from senior staff) to evaluate the agent's adherence to systematic debugging protocols.
Ingestion points: Hypothetical scenarios provided in the test-pressure-*.md files.
Boundary markers: Absent; the scenarios are presented as direct inputs for the agent to process.
Capability inventory: The skill uses local command execution (npm, git, security) to investigate bugs.
Sanitization: No specific sanitization or filtering is applied to the data processed from these scenarios.

systematic-debugging