The Agent Skills Directory

[Prompt Injection] (SAFE): No instructions designed to bypass agent constraints or safety filters were identified.
[Data Exposure & Exfiltration] (SAFE): No hardcoded credentials, sensitive file access (like SSH keys), or suspicious network requests were found. The skill focuses on local file scaffolding.
[Indirect Prompt Injection] (LOW): The scaffold_test.py script is designed to ingest external source code to generate tests. This presents an attack surface where malicious instructions embedded in source code comments could influence the agent's output.
Ingestion points: scaffold_test.py via the --source parameter.
Boundary markers: None specified in the documentation or usage examples.
Capability inventory: File generation and command execution via uv run.
Sanitization: No mention of sanitization for input source files.
[Command Execution] (SAFE): The skill provides usage examples for uv run to execute Python scripts. This behavior is consistent with the stated purpose of a developer scaffolding toolkit.
[Metadata Poisoning] (SAFE): The skill metadata correctly identifies its version, purpose, and the migration of its scripts to a unified builder skill.

moai-toolkit-codegen