The Agent Skills Directory

[SAFE] (SAFE): The skill is entirely instructional, providing markdown-formatted patterns for agent coordination.
[COMMAND_EXECUTION] (SAFE): No shell commands or subprocess calls are present in the documentation or metadata.
[EXTERNAL_DOWNLOADS] (SAFE): The skill does not reference or download any external packages or scripts.
[PROMPT_INJECTION] (SAFE): The instructions are clear and benign, focusing on domain-specific code analysis (security, backend, frontend) without attempting to bypass safety filters or override core agent instructions.
[DATA_EXFILTRATION] (SAFE): Allowed tools are restricted to Read, Glob, and Grep, which are used for local codebase analysis. There are no network-capable tools or exfiltration patterns.
[INDIRECT_PROMPT_INJECTION] (INFO): While the skill encourages agents to read codebase content (a potential injection surface), the skill itself is a meta-instruction set and does not contain logic that processes untrusted data in a way that introduces new risks beyond the agent's baseline behavior.

parallel-agents