ai_instruction_formatting

What problem does it solve?

This Skill helps you organize any LLM-consumed content into a consistent pseudo-XML structure so the model can reliably interpret roles, policies, inputs, and output contracts instead of relying on ambiguous prose ordering.

Core Features & Use Cases

• Pseudo-XML tagging for instruction clarity: Wrap semantic concerns in dedicated snake_case tags like <role>, <policy>, <inputs>, and <output_contract>.
• Works across many host artifacts: Applies to SKILL.md files, agent definitions, command/rules documents, or any instruction snippet consumed at inference time.
• Supports multiple file shapes: Handles prose-only markdown, full XML instruction bodies, tutorials with xml fenced examples, and mixed-agent documents with multiple top-level wrappers.
• Mechanical validation via bundled linter: Includes a deterministic linter that enforces structural correctness (naming, balancing, depth limits, and tag-structure constraints).

Quick Start

Ask an AI to format your existing system prompt or rule document into pseudo-XML using tags for role, objective, policy, inputs, steps, examples, and an output contract, then validate the result with scripts/lint_pseudo_xml.py.