The Agent Skills Directory

[SAFE]: The skill is authored by the platform vendor (Amplitude) and is designed to interact with its own analytics API. All tool calls referenced (e.g., query_charts, get_feedback_insights, get_deployments) are platform-specific and used within their intended functional scope.\n- [INDIRECT_PROMPT_INJECTION]: The skill possesses a data ingestion surface by processing content from feedback insights and experiment descriptions which could theoretically contain malicious instructions. However, this is categorized as safe because the skill lacks high-risk capabilities, such as arbitrary command execution or external network access, that would allow for an exploit. Additionally, Phase 3 of the instructions mandates a validation and filtering step to assess data confidence and relevance.\n
Ingestion points: get_feedback_insights and get_experiments (Phase 2).\n
Boundary markers: Absent; data is ingested as raw strings without specific delimiters.\n
Capability inventory: Tool access is restricted to the Amplitude environment (search, get_dashboard, query_charts, query_dataset).\n
Sanitization: Absent, though logical vetting is performed in Phase 3.

weekly-brief