The Agent Skills Directory

[Prompt Injection] (SAFE): The skill contains no instructions designed to override AI safety constraints or hijack the system prompt. Instructions are focused on technical documentation.
[Data Exposure & Exfiltration] (SAFE): No credentials or sensitive file paths are present. The validation script and patterns do not perform network requests.
[Obfuscation] (SAFE): All content is presented in clear, human-readable text without encoded or hidden payloads.
[Unverifiable Dependencies] (SAFE): All referenced packages (Vitest, Jest, React Testing Library) are industry-standard tools.
[Indirect Prompt Injection] (LOW): The skill is a reference tool (skill_type: reference). While it processes a user-defined 'topic', it lacks dangerous capabilities like shell execution or network writing that could be leveraged by an attacker through untrusted input.
[Dynamic Execution] (SAFE): The Python validation script uses yaml.safe_load() correctly to prevent unsafe deserialization of the configuration file.

testing