The Agent Skills Directory

[SAFE]: No malicious patterns or security risks were identified. The skill performs its stated function of screening and scoring ideas using local workspace data.
[COMMAND_EXECUTION]: The skill runs a local Python script to process data. It handles file paths safely using the pathlib library and restricts its operations to the provided workspace directory. It uses dynamic path insertion to import internal tooling, which is a standard pattern for repository-internal dependencies.
[PROMPT_INJECTION]: No evidence of prompt injection or bypass instructions was found in the skill metadata or script code. While the skill processes user-provided idea directions, it treats them as data for scoring rather than executable instructions.

idea-screener