test-skill

This skill serves as a dedicated testing fixture within the agnix ecosystem, designed specifically to validate the integrity of AI agent configuration files. It acts as a controlled environment for developers to trigger and observe rule-based diagnostic alerts, ensuring that agent-specific instructions and metadata remain compliant with platform-specific standards. By simulating real-world configuration errors—such as unsupported fields or invalid formatting—this fixture allows engineers to verify that their linting pipelines effectively catch potential issues before they cause silent failures in agent invocation or runtime behavior.

• Validates SKILL.md and agent metadata against cross-platform requirements including GitHub Copilot, Claude Code, and Cursor specifications.

• Demonstrates diagnostic triggers for CP-SK-001 and related rule-based error patterns.

• Facilitates regression testing for automated fixers and configuration linters.

• Monitors compliance for agent-driven workflows, ensuring configurations adhere to semantic guidelines.

• Supports the broader agnix suite by providing a benchmark for rule efficacy and parser accuracy.

• Intended for developers building, maintaining, or customizing AI-powered coding assistants and autonomous agents.

• Best utilized during the development phase of agent skill creation or when updating configuration schemas for new AI tools.

• Input: YAML or Markdown-based agent skill files; Output: Linting diagnostic reports containing error codes, line numbers, and actionable remediation advice.

• Constraints: Primarily focused on validation logic; must be used in conjunction with the agnix CLI or LSP server to generate meaningful output. Does not perform actual agent task execution but verifies the setup required for successful invocation. Ensure that target environments (e.g., Claude Code, Copilot, Kiro) are specified in the linting configuration to receive context-aware feedback.