Python track: exercise 08 doesn't mention re-running pip install after updating requirements.txt for local testing

開啟 exercises/Python/08-deploy-agent-to-cf.md，查看第 255-282 行的「Update requirements.txt」區段。新增本機安裝步驟，並註明在本機測試前需要執行此步驟，而 Cloud Foundry 會在部署期間安裝 requirements。確認轉譯後的練習清楚地區分本機測試與 cf push。

Summary

The "Update requirements.txt" section (08-deploy-agent-to-cf.md:255-282) instructs learners to add crewai[a2a] and a2a-sdk[http-server] to requirements.txt, then moves straight into creating manifest.yml for Cloud Foundry deployment. There's no pip install -r requirements.txt (or equivalent) command shown anywhere in the exercise.

Effect

This works fine for the actual CF deployment, since CF's Python buildpack automatically runs pip install -r requirements.txt during cf push — the deployed app gets the new dependencies without any local action needed.

However, if a learner wants to test the A2A server locally before deploying (a reasonable thing to want to do, and not explicitly discouraged anywhere), editing requirements.txt alone doesn't install anything into their existing venv. Running the server locally at that point fails with ModuleNotFoundError: No module named 'a2a', since the new packages were never actually installed — only added to a text file.

Suggested fix

Add an explicit step after the requirements.txt update, e.g.:
​bash pip install -r requirements.txt ​
with a short note that this is needed if testing locally before cf push (even though CF's buildpack handles it automatically during deployment).

Environment

Reproduced on a personal SAP AI Core tenant, Python (CrewAI + LiteLLM) track, testing the A2A server locally before Cloud Foundry deployment.