The Agent Skills Directory

[SAFE]: The skill's primary function is to generate configuration snippets and architectural designs based on user-provided parameters for distributed computing frameworks. Analysis across all 10 threat categories revealed no malicious patterns or vulnerabilities.\n- [PROMPT_INJECTION]: Instructions focus on technical system design and do not contain override attempts, role-play injections, or requests to bypass safety filters.\n- [DATA_EXFILTRATION]: No hardcoded credentials or suspicious network operations were found. The skill suggests best practices such as using Kubernetes secrets and IAM roles for managing storage credentials.\n- [REMOTE_CODE_EXECUTION]: The skill provides code templates (PySpark, Flink, Polars) but does not include any commands to download or execute untrusted remote scripts (e.g., no curl | bash patterns).\n- [OBFUSCATION]: The reference files contain highly repetitive technical sections, which appear to be intended as informational filler or context enrichment for the AI, but no hidden URLs, Base64-encoded payloads, or homoglyph attacks were detected.\n- [DEPENDENCIES]: References to external libraries (RAPIDS, Polars, Spark) involve well-known, established open-source projects. No suspicious or unpinned dependencies were identified.

data-distributed-compute