The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes npx eslint and a local bash script to perform code analysis. The execution is restricted by the skill's tool definition to specific linting commands.
[EXTERNAL_DOWNLOADS]: The use of npx may trigger the download of the eslint package from the official npm registry. As a well-known service and the intended tool for the task, this is documented neutrally as a safe operation.
[PROMPT_INJECTION]: A surface for indirect prompt injection exists because the agent reads and processes untrusted project files. 1. Ingestion point: Source code files read during the linting process. 2. Boundary markers: No specific markers are used to delimit code from instructions. 3. Capability inventory: The agent has Bash execution and File Write permissions. 4. Sanitization: Content is not sanitized before analysis.

eslint-fix