The Agent Skills Directory

[COMMAND_EXECUTION]: The skill executes a bundled script, scripts/attempt.mjs, to manage the state of tutorial exercises. Technical review of the script confirms it is used for administrative tasks such as computing folder digests and managing attempt logs in the local .re0/tutorial directory.
[DATA_EXPOSURE]: The tutorial script reads local file content to generate integrity hashes. This is a local-only operation used to detect version changes in dependencies. No network requests, API calls, or exfiltration patterns were found in the instructions or the supporting JavaScript code.
[INDIRECT_PROMPT_INJECTION]: The skill ingest data from the local codebase and learner submissions for grading purposes. The risk of malicious instructions in these inputs is mitigated by the disable-model-invocation: true setting in the frontmatter, which prevents the agent from performing unintended actions even if instructions are encountered in processed data.

re0-tutorial