The Agent Skills Directory

[SAFE]: The skill consists entirely of Markdown documentation and structural templates. It contains no executable code, scripts (e.g., .py, .js, .sh), or binaries that could perform unauthorized actions.\n- [SAFE]: Tool access is explicitly restricted in the YAML frontmatter to Read, Glob, and Grep. This limits the skill's capabilities to discovery and analysis, preventing file system modifications or network exfiltration.\n- [SAFE]: No obfuscation, hidden instructions, or credential exposure patterns were detected. The orchestration patterns described (Sequential, Parallel, Synthesis) follow standard development and auditing workflows.\n- [SAFE]: The skill does not define external dependencies or remote code execution vectors. All referenced agents (e.g., security-auditor, test-engineer) are conceptual roles within the platform's native orchestration system.

parallel-agents