The Agent Skills Directory

[SAFE]: The skill's instructions and prompts are designed for analytical tasks without requesting sensitive permissions or performing dangerous operations.
[INDIRECT_PROMPT_INJECTION]: The skill is designed to process external data sources such as historical traffic and resource metrics, which constitutes an indirect prompt injection surface. However, this is mitigated by the skill's specific design and lack of dangerous capabilities.
Ingestion points: User-supplied traffic logs, metrics, and business forecasts as defined in prompts/capacity-planning-analysis.md.
Boundary markers: The skill enforces the use of specific section headers (Working Assumptions, Open Questions, evidence) to isolate different types of information.
Capability inventory: The skill is limited to generating analytical artifacts and does not perform network operations, file system writes, or arbitrary command execution during its primary operation.
Sanitization: Instructions include auditing input for credibility and recency, as well as using masked data for production or privacy-sensitive work.

capacity-planning-analysis