The Agent Skills Directory

[SAFE]: The skill is designed for creative content generation, providing templates and reference files for novel planning.
[SAFE]: Data storage is localized to the project's work directory, specifically within the .sumeru/outline/ folder for JSON and Markdown outputs.
[SAFE]: The parallel processing logic using sub-agents is confined to text generation tasks and adheres to strict task-splitting rules.
[SAFE]: An indirect prompt injection surface was analyzed: user-provided novel topics are interpolated into templates. This is considered safe as the skill's capabilities are limited to generating creative writing data and do not involve network exfiltration or system-level command execution.
Ingestion points: User-provided story prompts in SKILL.md.
Boundary markers: None identified in the provided templates.
Capability inventory: Spawns sub-agents for task parallelization and writes files to the local directory.
Sanitization: Features an automated compliance check for real-world names to prevent legal infringement.

sumeru-outline