The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill uses npx ai-devkit@latest in its workflow, which fetches and executes the latest version of the package from the npm registry at runtime.
[COMMAND_EXECUTION]: The instructions provide command templates that interpolate user-provided issue descriptions (observed behavior) directly into shell arguments. This pattern is susceptible to command injection if the input contains shell-sensitive characters such as backticks or dollar signs.
[REMOTE_CODE_EXECUTION]: By using npx, the skill triggers the execution of remote code downloaded from a public registry, which introduces risks associated with package integrity and version stability.
[INDIRECT_PROMPT_INJECTION]: The skill provides a surface for processing untrusted data during the debugging lifecycle. Ingestion points: Data retrieved from the memory search tool and user-supplied reproduction steps. Boundary markers: The skill does not specify the use of delimiters to isolate external data from its core instructions. Capability inventory: The agent is authorized to execute shell commands and propose code modifications. Sanitization: The skill lacks instructions for sanitizing or validating user-provided strings before they are used in command-line tools.

structured-debug