The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides instructions for analyzing tracebacks and code, which serves as a potential ingestion point for malicious external content.\n
Ingestion points: Tracebacks and source code are read into the agent's context using the Read and Bash tools as part of the debugging workflow.\n
Boundary markers: The skill does not provide delimiters or specific instructions to disregard embedded commands that might be present in analyzed error logs.\n
Capability inventory: The agent is authorized to use Write, Edit, and Bash tools, creating a risk if instructions embedded in data are inadvertently executed.\n
Sanitization: There are no explicit steps provided for sanitizing or filtering the content of tracebacks before analysis.

python-debugger