The Agent Skills Directory

[SAFE]: The skill functions as an educational orchestrator, managing a 'teaching workspace' by reading from and writing to local files like MISSION.md, RESOURCES.md, and learning-records. This local-first approach minimizes external security risks.- [PROMPT_INJECTION]: The skill possesses a surface for indirect prompt injection as it is instructed to follow goals and insights defined in user-editable workspace files. While this is inherent to its stateful teaching design, no malicious injection attempts were found in the provided templates. • Ingestion points: MISSION.md, RESOURCES.md, learning-records/*.md, and NOTES.md. • Boundary markers: Absent; the agent is not explicitly told to ignore instructions embedded within these files. • Capability inventory: The agent can read and write files and is suggested to use CLI commands to open lesson files. • Sanitization: No content validation is performed on the ingested markdown files.- [DATA_EXFILTRATION]: No exfiltration patterns were detected. The skill uses external URLs only as example citations in documentation, adhering to standard educational practices.

teach