AgentCapture/docker-compose.dev.yml at main · Tcotl/AgentCapture · GitHub

1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
 # docker-compose.dev.yml
 # Hot-reload dev override for AgentCapture.
 #
 # Use this together with docker-compose.yml to start the stack with live
 # source reload instead of rebuilding the image on every change. Reuses the
 # same `honeypot_data` named volume, so the SQLite DB and knowledge_base
 # persist across reloads.
 #
 # Usage (mirrors the deploy.sh flag style):
 #   docker compose --env-file .env.docker \
 #     -f docker-compose.yml -f docker-compose.dev.yml up
 #
 # Follow logs:
 #   docker compose --env-file .env.docker \
 #     -f docker-compose.yml -f docker-compose.dev.yml logs -f honeypot
 #
 # Stop:
 #   docker compose -f docker-compose.yml -f docker-compose.dev.yml down
 #
 # To go back to the production container, run `./deploy.sh` as usual.

 services:
   honeypot:
     container_name: agent-capture-honeypot-dev
     # `uvicorn --reload` spawns a child watcher; PYTHONDONTWRITEBYTECODE=1
     # (set in the image) keeps the bind-mounted source free of stray .pyc
     # writes, but we still need root to drop the editable-install .pth into
     # site-packages and to bind ports below 1024 if you ever configure them.
     user: root
     environment:
       APP_ENV: development
     healthcheck:
       # Allow extra time for the one-time `pip install -e .` at first boot.
       start_period: 45s
     volumes:
       # Hot reload source mount: edit on host, uvicorn picks it up.
       - ./app:/app/app
       # Re-evaluated by `pip install -e .` on each container start, so dep
       # changes are picked up by restarting the container (not by --reload).
       - ./pyproject.toml:/app/pyproject.toml:ro
     command:
       - sh
       - -c
       - "pip install -e . >/tmp/pip.log 2>&1 && exec uvicorn app.main:app --host 0.0.0.0 --port $${PORT:-4877} --reload --reload-dir /app/app"