The Agent Skills Directory

[SAFE]: The skill operates as a structural framework for multi-agent coordination, relying on the platform's native Agent tool to dispatch tasks to a predefined catalog of 42 local agents in the .claude/agents/ directory. No external logic or third-party agent providers are utilized.
[SAFE]: The requested tools (Read, Glob, Grep) are strictly limited to read-only operations for codebase analysis, which is consistent with the skill's stated purpose and adheres to the principle of least privilege.
[SAFE]: No obfuscation techniques, base64-encoded strings, or hidden URLs were identified within the skill instructions or reference files. The documentation is transparent and uses standard markdown formatting.
[SAFE]: There are no external dependencies, package installations, or remote script execution patterns. All references point to local workspace documentation and configuration files that manage agent behavior.
[SAFE]: The orchestration patterns and synthesis protocols are defensive in nature, providing a structured way to evaluate code for security vulnerabilities, architecture quality, and framework correctness without introducing new risks.

parallel-agents