feat: add configurable settings and Qdrant shutdown

[Teagan42]

What

• add Settings config model
• allow injecting AsyncQdrantClient into PlexServer and close it on shutdown
• drive model names and Qdrant options from settings
• ensure tests and server shutdown close the client

Why

• enables dependency injection for easier testing
• avoids leaking Qdrant connections
• centralizes environment configuration

Affects

• mcp_plex.server
• mcp_plex.config
• tests for server and CLI

Testing

• uv run ruff check .
• uv run pytest

Documentation

• none

---

https://chatgpt.com/codex/tasks/task_e_68c64efc6df88328ad2ee1c5b8ee15f6

[github-actions]

Coverage Report

File	Stmts	Miss	Cover	Missing
mcp_plex
cache.py	36	4	89%	22, 50–52
loader.py	197	10	95%	32–34, 331, 334–338, 663
server.py	298	22	93%	59–60, 102–103, 226, 454, 458, 480–486, 518, 528, 559–560, 577–578, 601, 604–605, 663
TOTAL	651	36	94%

Tests	Skipped	Failures	Errors	Time
29	0 💤	0 ❌	0 🔥	31.983s ⏱️

[chatgpt-codex-connector]

Codex Review: Here are some suggestions.

Reply with @codex fix comments to fix any unresolved comments.

About Codex in GitHub

Your team has set up Codex to review pull requests in this repo. Reviews are triggered when you open a pull request for review, mark a draft as ready, or comment "@codex review". If Codex has suggestions, it will comment; otherwise it will react with 👍.

Codex can also answer questions or update the PR. Try commenting "@codex fix this CI failure" or "@codex address that feedback".

[chatgpt-codex-connector]

[P1] Declare pydantic-settings dependency

The new configuration module now imports pydantic_settings.BaseSettings to define Settings, but pyproject.toml still omits pydantic-settings from the project dependencies (only the version number was bumped). In a fresh installation the library will raise ModuleNotFoundError as soon as mcp_plex.config is imported. Add pydantic-settings to the dependency list so the package can start successfully.

Useful? React with 👍 / 👎.