The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill provides functionality for the agent to bind and process data from external remote sources, which is a potential surface for indirect prompt injection if untrusted data is processed.
Ingestion points: Data ingestion occurs through PivotViewDataSourceSettings.DataSource and remote data adaptors like WebApiAdaptor via SfDataManager.Url, as documented in SKILL.md and references/data-binding.md.
Boundary markers: The documentation explicitly instructs users and agents to use authenticated endpoints and secure configuration, and provides architectural patterns to isolate data processing from the client.
Capability inventory: The component performs data aggregation, layout rendering, and can trigger network requests to endpoints defined in the configuration. It does not perform arbitrary system command execution.
Sanitization: The skill includes a dedicated 'Security Warning: Data Source Validation' section in SKILL.md and references/data-binding.md, which mandates validation and sanitization of all data received from external sources before binding.
[EXTERNAL_DOWNLOADS]: The skill references official NuGet packages and GitHub repositories for its operations.
Evidence: The documentation points to Syncfusion.Blazor.PivotTable, Syncfusion.Blazor.AI, and examples hosted on the github.com/syncfusion organization. These are vendor-owned resources used for the skill's primary purpose and are considered safe.
[SAFE]: The skill provides explicit 'Best Practice' sections teaching the agent how to avoid security pitfalls such as hardcoded credentials and SQL injection by using IConfiguration and parameterized queries in server-side services.

syncfusion-blazor-pivot-table