XML Prompting & Semantic Anchors

Why XML works

Claude was trained on massive corpora that include XML. When you structure a prompt with tags, you create unambiguous delimiters that separate the instruction, context, reference code, and constraints. The result: Claude interprets each section exactly in the role you assign it, without having to infer what is instruction and what is context.

XML structure is relevant for complex or reusable tasks. For a one-line fix, it is unnecessary noise.

Basic structure

<instruction>
Extract the validation logic into a separate module
</instruction>

<context>
Express service, Node.js 20, no external dependencies
</context>

<code_example>
// Existing pattern to preserve
function validateEmail(email: string): boolean {
  return /^[^\s@]+@[^\s@]+\.[^\s@]+$/.test(email)
}
</code_example>

<constraints>
- Do not modify public function signatures
- Maintain compatibility with existing tests
</constraints>

<output>
New file src/validation/index.ts with exports
</output>

Reusable tag catalog

Semantic Anchors in code

Anchors are comment markers that allow Claude to reference specific areas of code without reading entire files.

// ANCHOR: auth-flow
export async function authenticate(token: string) {
  // ...
}
// END-ANCHOR: auth-flow

// ANCHOR: token-validation
function validateJWT(token: string): Claims {
  // ...
}
// END-ANCHOR: token-validation

In your prompt: “Modify only the ANCHOR: auth-flow section to add expired token handling.” Claude knows exactly where to intervene without reading the entire file.

When to use XML

Integration with CLAUDE.md

You can define the project XML conventions in CLAUDE.md to standardize team prompts:

## XML Prompt Conventions
Project tags: <api_design>, <accessibility>, <perf_budget>
For features: always include <context> and <constraints>

All team members then use the same tag vocabulary, making shared prompts (in .claude/commands/) consistent and understandable.