The Agent Skills Directory

[DATA_EXPOSURE_AND_EXFILTRATION]: The skill instructions provide examples for gathering diagnostic evidence, such as inspecting environment variables and keychain identities. These actions are context-bound to troubleshooting build and signing pipeline failures.
[REMOTE_CODE_EXECUTION]: The skill includes a utility script (find-polluter.sh) that iterates through and executes test files using npm test to isolate state pollution issues in test suites.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill requires the agent to analyze application logs and stack traces provided by the user during the investigation phase.
Boundary markers: None specified in the instructions.
Capability inventory: The skill utilizes shell execution for diagnostics and test running.
Sanitization: No explicit sanitization of ingested log data is described.

systematic-debugging