The Agent Skills Directory

[SAFE]: No malicious patterns or security vulnerabilities were detected. The skill is instructional and defines patterns for agent coordination.
[PROMPT_INJECTION]: Analyzed for indirect prompt injection surfaces. Ingestion points include Read, Glob, and Grep tools used for codebase mapping (SKILL.md). No explicit boundary markers or sanitization logic are defined to mitigate malicious content in processed files. However, the skill itself contains no malicious instructions or bypass attempts.
[DATA_EXFILTRATION]: The skill uses Read, Glob, and Grep tools for local codebase analysis. No external network requests or sensitive data harvesting patterns were identified.
[REMOTE_CODE_EXECUTION]: No external dependencies are fetched or executed at runtime. The orchestration logic relies on native agent capabilities within the host environment.

parallel-agents