us-ela-standards
Pass
Audited by Gen Agent Trust Hub on Oct 6, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill's research workflow instructs AI agents to visit external government websites to verify and update standards data. This ingestion of data from uncontrolled external domains represents a potential surface for indirect prompt injection attacks.
- Ingestion points: Official state education agency websites and document URLs provided in the research notes across the
references/states/directory and inreferences/frameworks.md. - Boundary markers: While the skill uses structured TSV files and Markdown headers internally, it does not specify explicit delimiters or "ignore previous instructions" warnings to be used when the agent processes data fetched from external URLs.
- Capability inventory: The skill provides Python scripts for data manipulation (
ccss-xml.py,ccss-merge.py) and explicit instructions inSKILL.md(Workflow C and D) for agents to update local repository files based on external research. - Sanitization: No explicit validation or filtering mechanisms are described for sanitizing content retrieved from external sites before it is integrated into the project's knowledge base or research files.
Audit Metadata