The Agent Skills Directory

[COMMAND_EXECUTION]: The skill's core functionality relies on executing a wide range of shell commands via the ddev CLI tool, including database management, service configuration, and container SSH access.
[EXTERNAL_DOWNLOADS]: The skill recommends installing external packages and add-ons from well-known registries, specifically chrome-devtools-mcp via npx and DDEV-managed add-ons using ddev add-on get.
[DATA_EXFILTRATION]: The ddev share command enables the intentional exposure of the local development environment to the public internet using third-party services like ngrok or cloudflared, which requires user awareness regarding data privacy.
[INDIRECT_PROMPT_INJECTION]: The integration of the Chrome DevTools MCP creates an attack surface where the agent ingests untrusted data from web pages (DOM, logs, network traffic).
Ingestion points: Browser data retrieved via the chrome-devtools-mcp server from active browser tabs.
Boundary markers: The skill does not provide specific boundary markers or instructions to the agent to ignore potentially malicious embedded content in the browser pages it inspects.
Capability inventory: Extensive shell execution capabilities (ddev), database access (ddev mysql, ddev import-db), and project file configuration.
Sanitization: No explicit sanitization or validation of browser-retrieved data is described before it enters the agent's context.
[DYNAMIC_EXECUTION]: The skill describes how to create and maintain custom shell scripts within the .ddev/commands/ directory, which are executed as commands within the DDEV environment.

ddev