headroom/docker-compose.yml
Tejas Chopra 481e0b83d5
fix(docker): publish compose ports on loopback only (#3061)
## Description

`docker compose up -d` published every service on `0.0.0.0`, and none of
the three authenticates an inbound caller by default:

| port | service | default auth |
|---|---|---|
| 8787 | proxy | `/v1/*` data plane open unless `HEADROOM_PROXY_TOKEN`
is set |
| 6333/6334 | Qdrant | **none at all** — holds embeddings derived from
prompts |
| 7474/7687 | Neo4j | `NEO4J_AUTH` falls back to `neo4j/devpassword`,
published in this file |

So the shipped default handed any peer on the surrounding network a
relay through the proxy plus direct read/write on the vector and graph
stores built from the operator's own prompt content. The proxy already
warns about exactly this shape at `headroom/proxy/server.py:3289` — the
compose file just never took its own advice.

## Type of Change

- [x] Bug fix (non-breaking change that fixes an issue)
- [ ] New feature (non-breaking change that adds functionality)
- [x] Breaking change (fix or feature that would cause existing
functionality to change)
- [ ] Documentation update

## Changes Made

- Pinned all five published ports to `127.0.0.1`.
- Documented in the file header how to expose the proxy deliberately,
pairing the port override with `HEADROOM_PROXY_TOKEN` rather than
leaving that implicit.
- Added a commented `HEADROOM_PROXY_TOKEN` entry to the proxy service
environment.
- Added a regression test asserting every published port names a
loopback host IP.

## Testing

- [x] Unit tests pass
- [x] Linting passes (ruff check + format)
- [ ] Type checking passes — N/A (YAML + test only)
- [x] New tests added for new functionality

### Test Output

```text
$ .venv/bin/python -m pytest tests/test_docker_compose_persistence.py -q
3 passed in 0.14s

$ docker compose -f docker-compose.yml config    # validates
headroom-proxy   host_ip=127.0.0.1  published=8787 -> 8787
neo4j            host_ip=127.0.0.1  published=7474 -> 7474
neo4j            host_ip=127.0.0.1  published=7687 -> 7687
qdrant           host_ip=127.0.0.1  published=6333 -> 6333
qdrant           host_ip=127.0.0.1  published=6334 -> 6334
```

Against the parent commit:

```text
FAILED test_top_level_compose_publishes_only_to_loopback
E   AssertionError: headroom-proxy: port '8787:8787' publishes on all interfaces
```

## Real Behavior Proof

- Environment: macOS 15 (darwin 25.4.0), Docker Compose v2 available
locally.
- Exact command / steps: `docker compose -f docker-compose.yml config
--format json` before and after, comparing the resolved `host_ip` on
every published port.
- Observed result: before, no port carried a `host_ip` (Docker binds
`0.0.0.0`); after, all five resolve to `host_ip=127.0.0.1`. The compose
file still validates.
- Not tested: bringing the stack up and probing the ports from a second
machine on the LAN — the assertion is made against Docker's own resolved
configuration rather than a live two-host network.

## Runtime Rollout Safety

- Rollout-managed feature(s): none.
- Minimum rollout channel: N/A.
- Stable/default behavior changed: yes — the compose stack is no longer
reachable from other machines by default.
- Kill switch / disable path: override `ports:` in a
`docker-compose.override.yml`; the header documents this and pairs it
with `HEADROOM_PROXY_TOKEN`.
- Unsafe override required: none.
- Qualification impact: none.
- Rollback path: revert this commit.

## Review Readiness

- [x] I have performed a self-review
- [x] This PR is ready for human review

## Checklist

- [x] My code follows the project's style guidelines
- [x] I have performed a self-review of my code
- [x] I have commented my code, particularly in hard-to-understand areas
- [x] I have made corresponding changes to the documentation (the
compose header)
- [x] My changes generate no new warnings
- [x] I have added tests that prove my fix is effective
- [x] New and existing unit tests pass locally with my changes

## Additional Notes

**This is a deliberate breaking change for one workflow**: anyone
reaching the compose proxy from another machine will need to override
`ports:`. That is exactly the configuration that was unsafe, so it
should break loudly rather than silently. `http://localhost:8787` from
the host is unchanged, the container still listens on `0.0.0.0`
internally, and service-to-service traffic on the compose network is
unaffected.

Scope note: I fixed all three services rather than only the proxy.
Closing 8787 while leaving an unauthenticated Qdrant and a
default-password Neo4j published on `0.0.0.0` would not have improved
the security posture.

Co-authored-by: Tejas Chopra <tejas@Tejass-MacBook-Pro.local>
2026-08-16 19:05:29 -07:00

121 lines
5.5 KiB
YAML

# =============================================================================
# Headroom — full "memory stack" compose
# =============================================================================
# Brings up the Headroom proxy together with the two datastores it needs for
# semantic memory: Qdrant (vector search) and Neo4j (relationship graph).
#
# Quick start:
# 1. cp .env.example .env # then set a real NEO4J_AUTH before any non-local use
# 2. docker compose up -d
# 3. point your LLM client at http://localhost:8787 (proxy)
#
# Just want the proxy without the memory features? You can run the proxy image
# on its own (`docker run -p 8787:8787 ghcr.io/headroomlabs-ai/headroom`); the two
# database services below are only required for the memory/relevance features.
#
# Ports published on the host — all bound to 127.0.0.1 (this machine only):
# 8787 proxy (OpenAI-compatible endpoint)
# 6333 Qdrant REST 6334 Qdrant gRPC
# 7474 Neo4j Browser 7687 Neo4j Bolt
#
# None of these three services authenticates inbound callers by default: the
# proxy's /v1/* data plane is open unless HEADROOM_PROXY_TOKEN is set, Qdrant
# has no API key, and Neo4j falls back to a published dev password. Publishing
# them on 0.0.0.0 therefore hands any peer on your network a relay through the
# proxy plus direct read/write on the embeddings and graph derived from your
# prompts. They are bound to loopback so that `docker compose up -d` is safe on
# a shared or untrusted network.
#
# To reach the proxy from another machine, publish it deliberately AND require
# a token — never one without the other:
# HEADROOM_PROXY_TOKEN=$(openssl rand -hex 32) # put this in .env
# ports: ["8787:8787"] # override in a compose override file
# =============================================================================
services:
# Headroom proxy — the OpenAI-compatible endpoint your client talks to.
# Built from the repo Dockerfile so it tracks your local checkout.
headroom-proxy:
build:
context: .
args:
HEADROOM_BUILD_VERSION: ${HEADROOM_BUILD_VERSION:-source-build}
# Bind to all interfaces inside the container so the published port is reachable.
command: ["--host", "0.0.0.0"]
environment:
- HEADROOM_HOST=0.0.0.0
- HOME=/home/nonroot
# Keep all Headroom read/write state on the named volume below.
- HEADROOM_WORKSPACE_DIR=/home/nonroot/.headroom
- HEADROOM_CONFIG_DIR=/home/nonroot/.headroom/config
# if you want to use a custom OpenAI-compatible API endpoint,
# uncomment and set the following line with the desired URL
# - OPENAI_TARGET_API_URL=https://api.x.ai
# Required before publishing this port beyond loopback: without it the
# /v1/* data plane accepts unauthenticated callers.
# - HEADROOM_PROXY_TOKEN=${HEADROOM_PROXY_TOKEN}
ports:
# Loopback-only. The container still listens on 0.0.0.0 (above) so the
# other compose services can reach it by name; this line controls only
# which host interfaces the port is published on.
- "127.0.0.1:8787:8787"
volumes:
- headroom_workspace:/home/nonroot/.headroom
# Readiness probe: the orchestrator polls /readyz so dependents and
# `docker compose up --wait` only see the proxy as healthy once it's serving.
healthcheck:
test: ["CMD", "curl", "--fail", "--silent", "http://127.0.0.1:8787/readyz"]
interval: 30s
timeout: 5s
retries: 3
start_period: 20s
# Start the datastores first. Note: this waits for the containers to start,
# not for them to be fully ready — the proxy retries its connections, so a
# brief "database not ready yet" window on first boot is expected.
depends_on:
- qdrant
- neo4j
# Vector database for semantic search.
# Stores embeddings so the proxy can retrieve semantically similar context.
qdrant:
image: qdrant/qdrant:v1.17.1
ports:
# Loopback-only: Qdrant runs unauthenticated here and holds embeddings
# derived from your prompts.
- "127.0.0.1:6333:6333" # REST API
- "127.0.0.1:6334:6334" # gRPC
# Named volume keeps the vector index across container restarts/recreates.
volumes:
- qdrant_data:/qdrant/storage
environment:
- QDRANT__SERVICE__GRPC_PORT=6334
# Graph database for relationships and multi-hop reasoning.
# Backs the memory features that traverse links between stored items.
neo4j:
image: neo4j:5.26
ports:
# Loopback-only: NEO4J_AUTH below defaults to a password published in
# this file, so an exposed Bolt port is an open database.
- "127.0.0.1:7474:7474" # HTTP (Browser)
- "127.0.0.1:7687:7687" # Bolt
# Named volume persists the graph data across container restarts/recreates.
volumes:
- neo4j_data:/data
environment:
# Credentials come from .env (NEO4J_AUTH=user/password). The default here
# is for LOCAL DEV ONLY — override it before exposing Neo4j anywhere.
- NEO4J_AUTH=${NEO4J_AUTH:-neo4j/devpassword}
# APOC: Neo4j's standard procedure library, needed by Headroom's queries.
- NEO4J_PLUGINS=["apoc"]
- NEO4J_apoc_export_file_enabled=true
- NEO4J_apoc_import_file_enabled=true
- NEO4J_apoc_import_file_use__neo4j__config=true
# Named volumes — managed by Docker, survive `docker compose down` (use
# `docker compose down -v` to delete the stored data as well).
volumes:
headroom_workspace: # persists dashboard savings/history, logs, config, memory state, session stats, and TOIN
qdrant_data:
neo4j_data: