The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill processes aggregated document audit findings from potentially untrusted sources. Maliciously crafted input data could attempt to influence the generated CSV output or the agent's operational logic during the reporting phase.
Ingestion points: The skill receives external "aggregated document audit findings" as its primary input (SKILL.md).
Boundary markers: There are no explicit delimiters or instructions to ignore embedded commands within the processed data.
Capability inventory: The skill is authorized to write files to the current working directory and access external documentation URLs.
Sanitization: The instructions include specific CSV formatting rules (Quote all text fields; escape internal quotes by doubling (" ")) that help mitigate CSV formula injection risks in the exported files.
[EXTERNAL_DOWNLOADS]: The skill references multiple external URLs for accessibility standards and remediation guidance.
Evidence: The skill points to official documentation from Microsoft (support.microsoft.com), Adobe (helpx.adobe.com, adobe.com), the W3C (w3.org), and the PDF Association (pdfa.org). These references are to well-known, trusted organizations and are used for documentation purposes.

document-csv-reporter