The Agent Skills Directory

[SAFE]: The skill consists of documentation and code templates intended to guide an AI agent in generating Flutter UI code. It does not include any executable scripts, remote code download commands, or attempts to access sensitive system files. The external source referenced, github.com/sarveshsea/memi, is the official repository for the project and is considered a safe vendor resource.
[PROMPT_INJECTION]: The skill processes external JSON component specifications to generate Dart source code, which represents a potential surface for indirect prompt injection. This is a characteristic of the skill's core functionality rather than a malicious pattern.
Ingestion points: Memoire component specifications in JSON format (as described in references/guide.md).
Boundary markers: The generator pipeline described in the guide includes a validation step using Zod schemas to verify the structure and integrity of input data.
Capability inventory: The skill is designed to emit static Dart source code, widget tests, and golden tests. It does not possess capabilities for arbitrary command execution or network exfiltration during the generation process.
Sanitization: The skill maps specific input fields directly to code templates, which serves to filter the input and prevent the injection of arbitrary instructions into the generated output.

flutter-gen