The Agent Skills Directory

[SAFE]: The skill focuses on standard mobile development workflows and does not contain any detected malicious instructions, obfuscated code, or dangerous command execution.
[PROMPT_INJECTION]: No override markers, bypass instructions, or attempts to extract system prompts were found. The core rules and execution protocol are focused on technical implementation and architectural integrity.
[DATA_EXFILTRATION]: There is no evidence of unauthorized sensitive file access or network exfiltration. The skill recommends best practices such as flutter_secure_storage for token management and utilizes Dio interceptors for standard API authentication.
[REMOTE_CODE_EXECUTION]: The skill does not perform any remote script downloads or dynamic execution of untrusted code. It references standard package managers (pub, npm) for dependency management within a well-defined tech stack.
[INDIRECT_PROMPT_INJECTION]: The skill has a defined surface for indirect injection as it reads existing codebase symbols and task requirements. However, it lacks dangerous capabilities like shell command execution or dynamic evaluation of that input, and focuses on generating structured code files based on templates.
Ingestion points: Reads project requirements and existing source code via get_symbols_overview and find_symbol tools.
Boundary markers: Not explicitly defined in the provided execution protocol.
Capability inventory: Code generation, file system writes (Step 3: Implement), and unit/widget testing.
Sanitization: Not explicitly mentioned, but the output is intended for developer review and build-time analysis.

mobile-agent