Your personal operating system for career transition.

Implement a Prompt Version Registry

When you run A/B evals or regression tests across prompt versions, you need a lightweight registry that ties eval results back to exactly the prompt text and metadata that generated them.

What to build

Implement PromptRegistry:

1. register(name, template, version, metadata=None) -> str — Returns a deterministic prompt_id as "{name}:{version}". Raises ValueError if that prompt_id already exists.

2. render(prompt_id, variables: dict) -> str — Substitutes {variable} placeholders in the template. Raises KeyError if a required variable is missing.

3. get(prompt_id) -> dict — Returns {"prompt_id", "name", "version", "template", "metadata"}. Raises KeyError if not found.

4. list_versions(name) -> list[str] — Returns prompt_ids for all versions of a named prompt, sorted lexicographically.

5. diff(prompt_id_a, prompt_id_b) -> dict — Returns {"same": bool, "template_a", "template_b"} showing the two templates side by side.

Constraints

• Standard library only. Templates use Python str.format_map style {variable} placeholders.