The Agent Skills Directory

[COMMAND_EXECUTION]: Provides instructions for executing development tools through the uv toolchain. The name 'pyright' is obscured using character insertion (p'y'ri'g'h't) throughout the document, which is a technique typically used to bypass automated text filters.
[EXTERNAL_DOWNLOADS]: References standard GitHub Actions and official Python documentation but includes a non-functional documentation link for 'ty'.
[REMOTE_CODE_EXECUTION]: Recommends the installation and execution of the 'ty' package from public registries. The skill incorrectly identifies this as an official Astral tool, which could lead to the execution of unintended third-party code.
[PROMPT_INJECTION]: The skill ingests local configuration files, creating a potential surface for indirect prompt injection. Ingestion points: Item 2 of the Agent Execution Checklist (reading pyproject.toml and lock files). Boundary markers: None provided in the instructions. Capability inventory: Execution of shell commands and modification of project files. Sanitization: No validation or sanitization is specified for the data read from these files.

modern-python