The Agent Skills Directory

[SAFE]: The skill instructions define clear workflows for engineering productivity analysis using the intended Harness MCP server. All tool invocations are scoped to the platform's API.
[SAFE]: The metadata and author information correctly identify the provider, and the skill's behavior is consistent with its stated purpose.
[SAFE]: While the skill ingests data from the 'harness_list' tool, it primarily performs analytical processing and lacks dangerous capabilities like arbitrary shell execution or file modification, making the indirect injection risk negligible.

sei-analytics