web-csv-reporter
Pass
Audited by Gen Agent Trust Hub on Sep 26, 2026
Risk Level: SAFEINDIRECT_PROMPT_INJECTIONEXTERNAL_DOWNLOADSDYNAMIC_EXECUTION
Full Analysis
- [INDIRECT_PROMPT_INJECTION]: The skill processes aggregated web audit findings which may originate from untrusted external websites, posing a potential indirect prompt injection surface.
- Ingestion points: The skill receives aggregated web audit findings as described in
SKILL.mdandreferences/csv-output-files.md. - Boundary markers: Explicitly defined in
SKILL.mdas a "read-only reporter" with instructions never to modify source documents or audit reports. - Capability inventory: Authorized to write CSV files and potentially scripts to the current working directory as noted in
SKILL.md. - Sanitization: Implements CSV-standard quoting and escaping for output fields to ensure data integrity during file generation.
- [EXTERNAL_DOWNLOADS]: The skill references accessibility documentation and help links from well-known sources, including the W3C and Microsoft's Accessibility Insights platform.
- [DYNAMIC_EXECUTION]: The skill may generate scripts if requested by a task to facilitate reporting, based on the structured audit findings and remediation plans defined in
references/csv-output-files.md.
Audit Metadata