The Agent Skills Directory

[EXTERNAL_DOWNLOADS] (SAFE): The skill references standard, well-known libraries in templates/package.json including @tanstack/react-table, @tanstack/react-query, and @tanstack/react-virtual. No suspicious or unknown external sources are utilized.
[COMMAND_EXECUTION] (SAFE): No risky shell commands or subprocess executions are present. The templates are standard React components for client-side and server-side table management.
[DATA_EXFILTRATION] (SAFE): There are no patterns suggesting data exfiltration. The network requests in the examples are directed to local API endpoints (/api/users) standard for web development.
[CREDENTIALS_UNSAFE] (SAFE): No hardcoded secrets, tokens, or private keys were found. The Cloudflare configuration uses placeholders like your-database-id.
[PROMPT_INJECTION] (SAFE): The skill instructions and README are strictly focused on technical implementation and do not contain override attempts or safety bypasses.
[INDIRECT_PROMPT_INJECTION] (SAFE): While the skill renders data from external APIs, it follows React's default escaping behavior.
Ingestion points: templates/server-paginated-table.tsx and templates/d1-database-example.tsx fetch data from /api/users.
Boundary markers: Standard UI boundaries; no specific LLM-prompt delimiters are used as the tool focuses on UI rendering.
Capability inventory: No OS-level or file-system capabilities.
Sanitization: The skill explicitly documents SQL injection prevention via whitelisting and prepared statements in references/server-side-patterns.md.

tanstack-table