fix(package): review pass on the PyPI package

- docsgpt api binds 127.0.0.1 by default, like gunicorn and uvicorn do;
  --host 0.0.0.0 exposes it. The docs say so.
- The embedded Milvus and LanceDB defaults derive from the data home, so
  they follow DOCSGPT_HOME like the faiss indexes and uploads do. A checkout
  run from its root and the Docker image resolve to the same paths as before.
- The docs and the pyproject comment describe the CPU torch install as two
  steps (torch and torchvision from the PyTorch CPU index first, then the
  docling extra): pip picks the highest version across indexes, so
  --extra-index-url only yields the CPU build while that index keeps pace
  with PyPI.
- AGENTS.md separates DOCSGPT_HOME (moves the data home) from
  DOCSGPT_ENV_FILE (selects the .env file); the docs example uses a password
  placeholder.

pyproject.toml

@@ -122,8 +122,9 @@ dependencies = [
 # attachment parsing, and read_document's `structured` output. Pulls torch and
 # transformers; with uv on Linux torch resolves from the CPU-only PyTorch index
 # (see [tool.uv.sources]) so the extra does not drag the CUDA stack in. pip
-# users on Linux get the same by adding
-# --extra-index-url https://download.pytorch.org/whl/cpu.
+# users on Linux install torch and torchvision from that index first
+# (--index-url https://download.pytorch.org/whl/cpu), then the extra; pip keeps
+# the torch it already has.
 docling = [
   "docling>=2.119.0,<3",
   "rapidocr>=3.9.2,<4",