The Agent Skills Directory

[SAFE]: All files provided in the skill are Markdown documents containing pedagogical principles, lesson templates, and workflow instructions for statistical teaching.
[PROMPT_INJECTION]: No instructions aimed at overriding agent constraints or bypassing safety guidelines were found in the instructional text.
[DATA_EXFILTRATION]: The skill does not contain any commands or code patterns associated with reading sensitive files (e.g., .ssh, .aws) or transmitting data to external domains.
[REMOTE_CODE_EXECUTION]: No remote script downloads (e.g., curl|bash) or external dependency installations are present.
[COMMAND_EXECUTION]: The skill does not execute shell commands or subprocesses. Mention of LaTeX macros and Mermaid diagrams are for documentation formatting purposes and do not involve runtime execution risks.
[CREDENTIALS_UNSAFE]: No hardcoded API keys, tokens, or secrets were detected in any of the skill files.
[INDIRECT_PROMPT_INJECTION]: The skill includes a 'handoff block' mapping in reference/workflow-and-handoff.md to ingest YAML data from other sources. However, because the skill has no dangerous capabilities like shell execution, network access, or file-write operations, this ingestion surface represents no significant risk.

statistical-pedagogy-framework