scaffold-exercises
Pass
Audited by Gen Agent Trust Hub on Sep 5, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONCOMMAND_EXECUTIONDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill is designed to read external source materials provided by the user, which introduces a risk of indirect prompt injection where malicious instructions could be embedded in the source data.\n
- Ingestion points: Reads user-specified lecture files (.tex, .qmd), research papers, and datasets as referenced in SKILL.md.\n
- Boundary markers: The skill instructions do not define delimiters or provide specific warnings to the agent to treat input content as untrusted.\n
- Capability inventory: The skill has access to powerful tools including Bash, Write, Read, Grep, and Glob as listed in the frontmatter.\n
- Sanitization: There is no mention of sanitizing or filtering the content retrieved from external sources before it is processed by the agent.\n- [COMMAND_EXECUTION]: The skill utilizes the Bash tool to execute code snippets, particularly for verifying coding solutions in R or Stata during the generation process.\n- [DYNAMIC_EXECUTION]: The skill generates and executes code snippets at runtime to ensure the solutions are correct and to capture the real output for inclusion in the solution key artifact.
Audit Metadata