The Agent Skills Directory

[COMMAND_EXECUTION]: The skill documentation in SKILL.md includes instructions for running a bundled Python script (scripts/annotate_code.py). This command is used to add educational markers to source code files for teaching purposes and operates on local files provided by the user.
[REMOTE_CODE_EXECUTION]: While references/common-mistakes.md contains examples of fetch() calls and network requests, these are exclusively fenced code blocks used as illustrative 'mistakes' or patterns for teaching JavaScript asynchronous programming. They are not instructions for the agent to execute code from the internet.
[CREDENTIALS_UNSAFE]: The references/common-mistakes.md file contains a reference to sqlite3.connect('db.db'). This is part of a teaching example demonstrating the 'Mixing Concerns' anti-pattern and does not involve hardcoded credentials or access to sensitive environment variables.
[EXTERNAL_DOWNLOADS]: The skill includes a skill-report.json file that references a GitHub repository for source context. This is a metadata reference for auditing and does not trigger unverified downloads during skill execution.

code-instructor