The Agent Skills Directory

[COMMAND_EXECUTION]: The skill is configured to use high-privilege tools such as 'shell' and 'git' to implement code changes and run tests within a target repository.
[SAFE]: No obfuscated content, hardcoded credentials, or suspicious remote download instructions were identified in the provided skill files.
[PROMPT_INJECTION]: The skill structure exhibits a surface for indirect prompt injection. Ingestion points: Repository content accessed via filesystem and git tools. Boundary markers: The workflow lacks delimiters or instructions to ignore embedded prompts in repository files. Capability inventory: The agent has access to shell, filesystem, and webfetch tools. Sanitization: No sanitization or validation logic is defined for data read from the repository.

dev-agent