Add Gemini and Codex CLI to Docker Compose

- Introduce gemini and codex services in compose.yaml for easier
  execution of AI CLI tools.
- Update README.md with instructions on how to build and run Gemini and
  Codex CLIs using Docker Compose.
- Clarify interactive CLI usage within Vim and via Docker Compose.

Author: jamm1985

README.md

@@ -113,7 +113,19 @@ docker compose build app
 docker compose run --rm app
 ```
 
-📓 Optional: Run JupyterLab
+- Optional: Run Codex or Gemini (see more examples below)
+
+```bash
+docker compose build codex
+docker compose run --rm codex
+```
+
+```bash
+docker compose build gemini
+docker compose run --rm gemini
+```
+
+- Optional: Run JupyterLab
 
 ```bash
 docker compose build jupyterlab
@@ -142,18 +154,29 @@ To open a terminal within Vim, you can use `:terminal` or `:vertical term` or
 `:tab terminal`. Once inside the terminal, you can invoke the CLI tools as
 usual.
 
+Alternatively, the AI CLI tools can be run outside Vim, in a separate terminal
+tab, window, or tmux pane. This is often preferable for longer interactive
+sessions or when reviewing large outputs. See, `codex` and `gemini` services in
+`compose.yaml`.
+
 ### Gemini CLI examples
 
 The Gemini CLI provides a conversational interface to interact with your
 codebase, allowing you to ask questions, refactor code, fix bugs, and add new
 features.
 
-Run interactively:
+Run interactively in Vim:
 
 ```bash
 gemini
 ```
 
+or using docker compose:
+
+```bash
+docker compose run --rm gemini
+```
+
 Read a file:
 
 ```bash
@@ -178,12 +201,18 @@ The Codex CLI (or similar code generation/analysis tools) can be used for
 automating code generation, understanding project structure, and suggesting
 improvements.
 
-Run interactively:
+Run interactively in Vim:
 
 ```bash
 codex
 ```
 
+or using docker compose:
+
+```bash
+docker compose run --rm codex 
+```
+
 Generate a new Python class (hypothetical):
 
 ```bash

compose.yaml

@@ -55,6 +55,32 @@ services:
       - type: bind
         source: .
         target: /application
+  gemini:
+    platform: linux/amd64
+    entrypoint: gemini
+    build:
+      target: vim-ide
+      args: *default-args
+    environment:
+      <<: *default-env
+      GEMINI_API_KEY: ${GEMINI_API_KEY}
+    volumes:
+      - type: bind
+        source: .
+        target: /application
+  codex:
+    platform: linux/amd64
+    entrypoint: codex
+    build:
+      target: vim-ide
+      args: *default-args
+    environment:
+      <<: *default-env
+      OPENAI_API_KEY: ${OPENAI_API_KEY}
+    volumes:
+      - type: bind
+        source: .
+        target: /application
   jupyterlab:
     platform: linux/amd64
     entrypoint: