The Agent Skills Directory

[EXTERNAL_DOWNLOADS]: The skill documentation suggests configuration for an MCP server that fetches the next-devtools-mcp package from the npm registry. It also references using @next/codemod and @next/playwright as part of the development and migration workflows.\n- [REMOTE_CODE_EXECUTION]: The recommended configuration for the Next.js devtools MCP server utilizes npx -y next-devtools-mcp@latest, which downloads and executes the latest version of the tool from a remote registry at runtime. Similar patterns are suggested for migration tasks using npx @next/codemod@canary.\n- [COMMAND_EXECUTION]: The instructions require the execution of several CLI tools for core functionality, including next typegen to generate route types, next build and next start for production verification, and agent-browser commands for runtime UI checks.\n- [INDIRECT_PROMPT_INJECTION]: The skill is designed to ingest and analyze untrusted project data which could be manipulated to influence the agent's behavior.\n
Ingestion points: Reads next.config.js, package-lock.json, and the source code tree (e.g., app/ or src/app/) to determine optimization strategies.\n
Boundary markers: None identified; the instructions do not specify delimiters or sanitization for the project files being processed.\n
Capability inventory: Includes listing MCP tools, fetching compilation logs, querying route metadata, and triggering builds.\n
Sanitization: There is no mention of validating or sanitizing the content of project files before the agent incorporates them into its reasoning.

nextjs