The Agent Skills Directory

[SAFE]: The skill performs legitimate development tasks consistent with its stated purpose. All observed behaviors, such as file modifications for fixing specs and running local analysis scripts, are transparently documented and part of the core functionality.\n- [COMMAND_EXECUTION]: The skill executes the spectral CLI tool for API linting and utilizes platform-specific commands (open, xdg-open, or start) to allow users to view generated assessment reports in their default browser. These executions are user-triggered and necessary for the tool's operation.\n- [SAFE]: The 'Fix Workflow' involves in-place editing of OpenAPI specification files to remediate identified issues. These edits are targeted and guided by rules defined in the skill's reference files, posing no risk to the execution environment beyond the intended file modifications.

api-design