The Agent Skills Directory

[External Script Execution]: The skill references an installation method for the Aspire CLI that involves fetching a shell script from https://aspire.dev/install.sh and piping it to bash. While this is a common pattern for developer tools, it involves executing remote code. The domain aspire.dev is the official project site for .NET Aspire, which is managed by the skill's author, aligning with the project's primary purpose.
[Command Execution]: The instructions direct the agent to invoke the aspire CLI for operations like init, new, and doctor. These commands are necessary for the skill to perform project scaffolding and health checks within the user's workspace.
[Indirect Prompt Injection Surface]: The skill determines its actions by scanning the workspace for specific files (e.g., .csproj, package.json, aspire.config.json) and allows the agent to derive project names from the environment. This represents a potential surface where maliciously crafted project structures or metadata could influence the arguments passed to CLI commands. The skill uses the --non-interactive flag as a mitigation against interactive execution risks.
Ingestion points: Project file structure and filename detection logic in SKILL.md.
Boundary markers: None explicitly defined for file content scanning.
Capability inventory: Shell command execution via the aspire CLI.
Sanitization: The skill relies on standard CLI argument parsing; no specific input sanitization for project-derived strings is documented.

aspire-init