Ratnesh Maurya — Software Development Engineer specialising in Go, Elixir, PostgreSQL, Kubernetes, and cloud-native backend systems. Currently building at Initializ.
I built this because every AI coding session starts from zero context.
You explain the same things each time "this is a Python project,
we use FastAPI, follow these patterns..." cursor-claude-personas is a library of 38 persona folders. Each one
contains .claude/ and .cursor/ configs with domain-specific rules
and skills. You copy a folder into your project root, reload your
editor, and the AI behaves like a specialist. For example, the security-engineer persona threat-models by default,
flags OWASP issues, and asks about blast radius before suggesting
fixes. The senior-python-developer persona enforces typing, knows
common FastAPI patterns, and structures code differently than a
generic assistant would. No magic. No prompt engineering. Just config files that both Claude
Code and Cursor read natively. MIT licensed: https://github.com/ratnesh-maury...
