The Agent Skills Directory

[PROMPT_INJECTION]: The skill instructions contain standard steering constraints and role-play instructions consistent with a senior architect persona. No attempts to bypass safety filters or override system instructions were detected.
[DATA_EXFILTRATION]: No sensitive file paths or credentials were identified. Network operations described in the reference guides are standard architectural patterns (e.g., gRPC, REST calls between services) and do not target suspicious domains or attempt to exfiltrate user data.
[REMOTE_CODE_EXECUTION]: The skill does not contain any patterns for downloading and executing remote scripts. All code examples are provided as static text for educational or implementation guidance.
[OBFUSCATION]: A thorough check for Base64, zero-width characters, homoglyphs, and hidden text patterns across all 6 files yielded no findings.
[DYNAMIC_EXECUTION]: No usage of unsafe deserialization (e.g., pickle) or dynamic code evaluation (e.g., eval/exec) was found in the provided code examples.
[INDIRECT_PROMPT_INJECTION]: While the skill is designed to ingest and analyze user-provided system requirements, it does not demonstrate capabilities that would allow for high-risk exploitation if malicious instructions were embedded in that data. The skill's primary output is architectural documentation and diagrams.

microservices-architect