The Agent Skills Directory

[COMMAND_EXECUTION]: The skill uses specialized tools to execute CLI commands (very_good create and very_good packages get) for project scaffolding and dependency management. These operations are core to the skill's purpose and are handled through the Very Good CLI MCP server.
[INDIRECT_PROMPT_INJECTION]: The skill processes user-supplied input to define project parameters, which represents a managed attack surface.
Ingestion points: User-provided project name, organization name, and template selection gathered via prompts in SKILL.md.
Boundary markers: None explicitly defined between user input and tool parameters.
Capability inventory: Filesystem modifications and network access for package resolution via mcp__very-good-cli__create and mcp__very-good-cli__packages_get in SKILL.md.
Sanitization: Instructions in SKILL.md require the agent to automatically convert dashes in project names to underscores to comply with Dart naming rules.
[SAFE]: No malicious obfuscation, credential harvesting, or unauthorized persistence mechanisms were detected. The skill's behavior aligns with the primary purpose provided by the vendor.

create-project