The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: Recommends installing the pyscn package from the Python Package Index (PyPI) using pip install or running it dynamically via uvx.
[COMMAND_EXECUTION]: Instructs the agent to execute the pyscn CLI tool for comprehensive code analysis, including complexity, dead code, and dependency cycle detection.
[PROMPT_INJECTION]: The skill processes untrusted Python source code and instructs the agent to summarize the results, which introduces a surface for indirect prompt injection.
Ingestion points: Reads and analyzes files from user-specified local directories (e.g., src/).
Boundary markers: No specific delimiters or instructions are provided to the agent to ignore potentially malicious content embedded in the source code comments or strings being analyzed.
Capability inventory: Execution of shell commands via the CLI and reading generated JSON/HTML report files from the .pyscn/reports/ directory.
Sanitization: There is no mention of sanitizing or escaping the content extracted from the analysis reports before it is processed by the agent.

cli-analysis