The Agent Skills Directory

[SAFE]: The skill is a standard instructional guide for diverse ideation. It references several academic sources (arXiv papers) to ground its methodology in research.
[COMMAND_EXECUTION]: The skill mentions internal functions such as select_methods, rotation_plan, and score_idea from the tome and leyline namespaces. These appear to be legitimate platform-provided tools for managing the ideation workflow and evaluating outputs.
[DATA_EXFILTRATION]: No network operations or attempts to access sensitive local files (e.g., SSH keys, credentials) were found. The skill operates entirely within the reasoning and prompting context.
[PROMPT_INJECTION]: The instructions do not contain attempts to bypass safety filters or override system prompts. It uses standard instructional language to guide agent behavior towards diversity in generation.
[EXTERNAL_DOWNLOADS]: There are no remote script downloads or unverified package installations. The references to arXiv papers are for citation purposes and do not involve active network fetching by the agent.
[INDIRECT_PROMPT_INJECTION]: While the skill processes user-provided problems (untrusted data), it lacks high-risk capabilities like file system writes or network exfiltration that would make it a significant target for indirect injection. The use of a scoring framework (score_idea) provides a layer of validation for novelty claims.

ideate