The Agent Skills Directory

[SAFE]: The skill is designed for security auditing of Flutter applications. It includes search patterns for detecting hardcoded secrets and provides actionable guidelines for secure coding practices.
[COMMAND_EXECUTION]: The skill uses standard shell commands (grep) to search for sensitive patterns in local project directories (lib/, android/, ios/). These operations are consistent with its stated purpose of performing security audits.
[INDIRECT_PROMPT_INJECTION]:
Ingestion points: The skill audits local source code and configuration files within the lib/, android/, and ios/ directories.
Boundary markers: The skill does not explicitly define boundary markers to separate audited content from instructions, but its narrow focus on pattern matching reduces the likelihood of instruction execution from data.
Capability inventory: Capabilities are limited to file reading via grep and writing a security review document.
Sanitization: No automated sanitization of audited content is described, though the skill provides strict behavioral instructions for the agent to follow during the audit process.

flutter-security