run-tests

name	run-tests
description	Runs all unit tests (Jest) and Python evaluations (tests/eval/run_evals.py) for the AstroTyper project, then reports the results. Does NOT refactor or fix any failures unless explicitly asked.

Run Tests & Evaluations

This skill runs the full test suite for the AstroTyper project and reports the results.

Steps

1. Run Jest unit tests (TypeScript game engine tests):

   source ~/.nvm/nvm.sh && npx jest 2>&1
   

   Run this from the project root: /Users/eugenehuang/Desktop/type-your-adventure-antigravity

2. Run Python narrative evaluations (LLM-as-judge):

   ⚠️ The IDE sandbox blocks pyenv's libpython3.10.dylib, so this CANNOT be run by the agent.

   After reporting the Jest results, remind the user to run this command manually in their terminal:

   .venv/bin/python tests/eval/run_evals.py
   

   This requires GEMINI_API_KEY in .env.local.

3. Report results — After both commands complete, summarize:

   • Total test suites and individual tests passed/failed
   • Any specific test failure messages (assertion mismatches, errors)
   • Python evaluation pass/fail status
   • Do NOT refactor or fix anything. Only report the results. Wait for the user to explicitly request fixes.