The Agent Skills Directory

[SAFE]: The skill's primary function is to provide benign instructional guidance and project management checklists.- [SAFE]: File system access is limited to verifying the existence of specific directories and reading basic project configuration (.axhub/make/axhub.config.json), which is consistent with its stated management purpose.- [SAFE]: The skill recommends legitimate browser-based MCP tools (Chrome DevTools, Playwright) to improve communication efficiency but does not automate their installation or execution.- [SAFE]: A potential indirect prompt injection surface is noted where the skill reads project metadata from a configuration file; however, no sensitive capabilities are exposed to this data, and the behavior is considered safe within the project's context. (Ingestion: .axhub/make/axhub.config.json; Boundaries: absent; Capability inventory: none; Sanitization: absent).

project-guide