AI skills for Erlang
AI Coding Skills for Erlang
Browse the best AI coding skills for Erlang in 2026, OTP 27 features, Rebar3 builds, Common Test + eunit, Cowboy HTTP servers, and Mnesia / RocksDB integration.
Short answer
The best AI coding skills for Erlang are skills-hub's `code-review`, `unit-test` (Common Test + eunit + PropEr), `api-scaffold` (Cowboy + Plug-equivalent), `db-schema` (Mnesia / RocksDB), and `secure`. Portable across Claude Code, Cursor, Codex CLI, Copilot, and Cline.
Erlang in 2026 still powers telecom-grade messaging, WhatsApp scale, and increasingly LiveView / Phoenix backbones via the BEAM. The skills below detect your Erlang stack (OTP 27, Rebar3, Cowboy, Mnesia or RocksDB) and emit idiomatic OTP, proper gen_server / gen_statem trees, supervisor strategies, and let-it-crash discipline.
Top skills for Erlang
Frequently asked questions
Do AI skills follow OTP discipline?
Yes, modern skills emit gen_server callbacks, supervisor trees with one_for_one / rest_for_one / one_for_all strategies as appropriate, and don't add try/catch where 'let it crash' is the right answer.
Can AI generate Cowboy HTTP handlers?
Yes, the api-scaffold skill emits Cowboy handlers with proper init/2, handle/3, terminate/3, and websocket handlers where the route opts in.
What's the best AI tool for Erlang?
Claude Code handles OTP refactors well thanks to Erlang's representation in training data. Cursor with the Erlang LS extension gives the best inline experience.
Do AI skills handle Mnesia transactions?
Yes, the db-schema skill emits Mnesia table definitions with proper indices, mnesia:transaction wrappers, and matches the project's existing conventions for dirty vs transactional reads.