The Agent Skills Directory

[INDIRECT_PROMPT_INJECTION]: The skill ingests natural language requests from users to plan project architectures and coordinate specialist agents. This creates an attack surface where malicious instructions embedded in a user request could influence the agent's behavior during the planning or execution phases.
Ingestion points: User requests describing application features (processed in SKILL.md).
Boundary markers: No explicit prompt delimiters or 'ignore embedded instructions' warnings are specified for the interpolation of user requirements into the planning process in SKILL.md.
Capability inventory: The skill has access to Bash for command execution and Agent for delegating tasks to other agents as defined in SKILL.md frontmatter. Specialist tools and execution capabilities are detailed in various TEMPLATE.md files.
Sanitization: No explicit validation or filtering of user-provided content is mentioned before it is used to generate the project plan.
[COMMAND_EXECUTION]: The skill instructs the agent to execute various shell commands for project initialization, package management, and running development servers.
Evidence: Use of npx nuxi@latest init, npm install, and npx prisma init in templates/nuxt-app/TEMPLATE.md; and uvicorn app.main:app in templates/python-fastapi/TEMPLATE.md.
[EXTERNAL_DOWNLOADS]: The skill facilitates the download and installation of numerous third-party dependencies and framework scaffolds from established registries and official initialization tools.
Evidence: Commands such as npm install and npx create-next-app target well-known services and trusted organizations for project scaffolding (e.g., in templates/nextjs-fullstack/TEMPLATE.md).

app-builder