The Agent Skills Directory

[SAFE]: The skill manages Perses platform operations using official tools and well-documented procedures. Methodology for dashboard creation, migration, and deployment aligns with standard industry practices.
[SAFE]: External references and downloads target official Perses project repositories on GitHub, the official Helm chart repository, and the primary Docker Hub image. These are trusted sources for the intended functionality.
[SAFE]: Instruction sets for handling credentials (such as API tokens for Grafana or Perses) correctly advise against hardcoding and instead recommend using secure environment variables or platform-native secret management (e.g., GitHub Secrets).
[SAFE]: The configuration of Claude Code settings for MCP integration is a legitimate part of the skill's onboarding workflow and is clearly documented for the user.
[SAFE]: Resource definitions (CUE, JSON, YAML) are subject to structural validation via percli lint throughout all operational phases, reducing the risk of accidental misconfiguration.

perses