The Agent Skills Directory

[PROMPT_INJECTION]: The skill is susceptible to indirect prompt injection through the analysis of untrusted Flutter/Dart source code.
Ingestion points: Reads files within the lib/ directory and pubspec.yaml using shell utilities like grep and find via the workflow steps in SKILL.md.
Boundary markers: The skill lacks explicit instructions or delimiters for the agent to ignore instructions embedded within the source code being analyzed.
Capability inventory: The skill uses local read-only commands (find, grep, wc). It does not have network access or file-write permissions.
Sanitization: Content extracted from the audited codebase is not sanitized or escaped before being processed by the agent.
[COMMAND_EXECUTION]: The skill workflow relies on the agent executing shell commands (find, grep, wc) to perform the audit. While these are local and intended for read-only analysis, they constitute a primary operational capability of the skill.
[SAFE]: The skill operates entirely locally and does not perform any network requests, external data transfers, or sensitive file access (e.g., SSH keys or environment variables) outside of the intended project scope.
[SAFE]: No hardcoded credentials, persistence mechanisms, or obfuscated code patterns were detected in the skill files.

atomic-design-system-auditor