The Agent Skills Directory

[PROMPT_INJECTION]: An indirect prompt injection surface exists because the agent is designed to ingest and analyze external code files that could contain malicious instructions.
Ingestion points: External data enters the agent context through the use of the Read, Grep, and Glob tools during the code review process.
Boundary markers: The skill does not implement delimiters or specific instructions to the agent to treat the reviewed code as untrusted data or to ignore embedded instructions.
Capability inventory: The agent has access to a powerful toolset including Bash, Write, and Edit, which could be used to perform unauthorized actions if a prompt injection attack is successful.
Sanitization: There is no evidence of sanitization, filtering, or escaping logic applied to the content of the files before they are processed by the underlying model.

flutter-reviewer