The Agent Skills Directory

Prompt Injection (LOW): The skill possesses an indirect prompt injection surface as it reads local project files to inform its patterns. Ingestion points: Reads files like src/entities/*.py and pyproject.toml using the Read tool. Boundary markers: Absent; no instructions to ignore embedded commands are present in the skill instructions. Capability inventory: Uses filesystem reading tools (Read, Grep, Glob) which influence agent code generation. Sanitization: Absent; the skill does not filter or validate the content of the project files it reads.
Data Exposure & Exfiltration (SAFE): No evidence of hardcoded credentials, sensitive file access, or network communication patterns was detected.
Obfuscation (SAFE): No encoded content, hidden characters, or other obfuscation techniques are present.
Remote Code Execution (SAFE): No patterns for downloading or executing remote code or untrusted packages were found.
Dynamic Execution (SAFE): The skill provides static templates and does not perform runtime code generation or unsafe data deserialization.

pygame-patterns