The Agent Skills Directory

[COMMAND_EXECUTION]: The skill instructs the agent to use subprocess.run for validating JavaScript syntax via node --check and for running browser-based verification via Playwright (SKILL.md). Additionally, the start.sh template executes shell commands to manage Docker containers and health checks (references/start.sh).\n- [DYNAMIC_EXECUTION]: The skill's primary purpose involves generating Python and JavaScript code that is executed at runtime to serve the dashboard interface (SKILL.md).\n- [INDIRECT_PROMPT_INJECTION]: The skill fetching external market news and financial data creates an indirect prompt injection surface. Ingestion points: Yahoo Finance news and ticker data via MCP tools (SKILL.md). Boundary markers: Use of json.dumps for data embedding and a CSP-safe CSS foundation (SKILL.md, references/ui-components.md). Capability inventory: subprocess.run, docker, and playwright (SKILL.md, references/start.sh). Sanitization: Instructions for CSP-safe event listeners, warnings against innerHTML, and a grep validation check (SKILL.md).\n- [EXTERNAL_DOWNLOADS]: UI libraries (Chart.js, Plotly.js, Lightweight Charts) are fetched from well-known CDNs including jsDelivr, Plotly, and Unpkg (references/chart-patterns.md).\n- [PRIVILEGE_ESCALATION]: The start.sh script attempts to start the Docker daemon (dockerd) in the background, which is a high-privilege operation in a sandbox environment (references/start.sh).

interactive-dashboard