The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill automatically installs the external Python package 'desloppify' from a public registry (PyPI) using pip install if the tool is not found on the system path.
[COMMAND_EXECUTION]: The skill invokes several shell commands to manage its lifecycle, including package installation via pip, checking for command existence with which, and executing the core functionality through various desloppify subcommands (scan, next, resolve).
[PROMPT_INJECTION]: The skill exhibits an attack surface for indirect prompt injection as it ingests and processes content from an untrusted codebase.
Ingestion points: Reads all files within the target directory during the scan and next operations (SKILL.md).
Boundary markers: No delimiters or instructions are provided to the agent to disregard potentially malicious instructions embedded in the scanned code comments or documentation.
Capability inventory: The skill has the ability to install software, execute shell commands, and read/write files on the local system.
Sanitization: There is no evidence of sanitization or validation performed on the contents of the codebase before it is processed by the agent.

desloppify