The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes untrusted data such as user-provided bug symptoms, stack traces, and project codebases to perform debugging and root cause analysis.
Ingestion points: External data enters the context via user input (symptoms) and file content retrieved through Read and Grep tools.
Boundary markers: The skill does not define explicit delimiters or instructions to ignore embedded content within processed data, though it follows a rigid analytical structure.
Capability inventory: The agent is authorized to execute shell commands via Bash and perform file system operations using Read, Grep, and Glob.
Sanitization: The risk is partially mitigated by the mandatory scientific verification process (chain-of-verification) and the requirement for automated regression tests to confirm findings.

deep-debugging