Files
gitops-automation/docker-compose.yml
william a2c00f6f8b Fix claude-bot replying to Hermes's own messages, route Hermes through LiteLLM
Critical bug: claude-bot only ignored its OWN messages and text explicitly
addressed to @hermes — it did not ignore Hermes's own replies appearing in
the room. A single @hermes mention cascaded into claude-bot replying to
Hermes's thread messages ('Hermes says: ...'), which could itself cascade
further. Fixed with an OTHER_AGENT_USER_IDS allowlist of sender IDs to
always ignore, not just a text-prefix check.

Also point Hermes's model provider at the local litellm gateway
(OPENAI_BASE_URL/OPENAI_API_KEY) instead of OpenRouter directly, matching
claude-agent's own chat path — one place to hold the OpenRouter credential.
This does NOT grant Hermes access to the Claude subscription; that's an
Anthropic-side restriction unrelated to which proxy sits in front of it,
already proven earlier in this session.

Separately (not a code fix): the 'Billing or credits exhausted: HTTP 402'
error Hermes hit is real — the OpenRouter account currently has 0 credits.
2026-08-23 16:02:15 +00:00

171 lines
8.4 KiB
YAML

services:
# Traefik deliberately does NOT live in this stack — it's shared infra fronting
# Gitea/Portainer/Matrix/agent (see ~/traefik/docker-compose.yml, a separate,
# independently-managed stack). It used to be a service here, but a GitOps redeploy
# tears every service in a stack down before bringing them back up — and claude-agent's
# image pull goes through Traefik→Gitea's registry, so a self-hosted Traefik ends up
# briefly tearing down the very route its sibling service needs to come back up.
# Circular dependency, self-inflicted outage. Don't put Traefik back in this file.
matrix-homeserver:
image: ghcr.io/continuwuity/continuwuity:latest
container_name: matrix-homeserver
restart: unless-stopped
environment:
CONTINUWUITY_SERVER_NAME: ${MATRIX_SERVER_NAME}
CONTINUWUITY_DATABASE_PATH: /var/lib/continuwuity
CONTINUWUITY_ADDRESS: 0.0.0.0
CONTINUWUITY_PORT: 8008
# Private control-room bot only — no federation, no open registration.
# Registration is flipped on temporarily, once, to create the bot account
# (see README "First boot: Matrix bot account").
CONTINUWUITY_ALLOW_FEDERATION: "false"
CONTINUWUITY_ALLOW_REGISTRATION: ${MATRIX_ALLOW_REGISTRATION:-false}
# CONTINUWUITY_REGISTRATION_TOKEN is deliberately NOT set here: Continuwuity treats
# a present-but-empty value as a config error (fails to start), so it can only be
# added here temporarily (with a real value) when you need to register a new user,
# then removed again — see README "Adding another Matrix user".
volumes:
- matrix_data:/var/lib/continuwuity
networks:
- web
labels:
- "traefik.enable=true"
- "traefik.http.routers.matrix.rule=Host(`${MATRIX_SERVER_NAME}`)"
- "traefik.http.routers.matrix.entrypoints=websecure"
- "traefik.http.routers.matrix.tls.certresolver=letsencrypt"
- "traefik.http.services.matrix.loadbalancer.server.port=8008"
litellm:
# Pinned deliberately, not :latest or :main-latest — litellm==1.82.7/1.82.8 on PyPI
# were compromised with credential-stealing malware in March 2026 (fixed within the
# hour, but a floating tag could still land on a bad release in the future). v1.98.0
# verified clean as of this writing.
image: ghcr.io/berriai/litellm:v1.98.0
container_name: litellm
restart: unless-stopped
environment:
OPENROUTER_API_KEY: ${OPENROUTER_API_KEY}
LITELLM_MASTER_KEY: ${LITELLM_MASTER_KEY}
volumes:
# Absolute host path, NOT a repo-relative one — Portainer's git-stack deploy clones
# into its own directory (/data/compose/N/) whose checkout doesn't reliably persist
# for the container's runtime (see the act_runner config comment below for the same
# failure mode). An absolute path on the actual host filesystem always resolves the
# same way regardless of which tool ran `docker compose up`. Keep this local clone
# (/home/william/gitops-automation) pulled to latest when the config changes.
- /home/william/gitops-automation/litellm-config.yaml:/app/config.yaml:ro
command: ["--config", "/app/config.yaml", "--port", "4000"]
networks:
- web
# Internal only — no Traefik labels. No reason to expose an LLM gateway holding a
# master key and OAuth-forwarding config to the public internet.
hermes:
# Pinned to a specific dated release, not :latest — same rationale as litellm above.
image: nousresearch/hermes-agent:v2026.8.19
container_name: hermes
restart: unless-stopped
environment:
HERMES_UID: "1000"
HERMES_GID: "1000"
# Internal container address, not the public HTTPS one — same docker network as
# matrix-homeserver, no reason to round-trip through Traefik/TLS for this.
MATRIX_HOMESERVER: http://matrix-homeserver:8008
MATRIX_ACCESS_TOKEN: ${HERMES_MATRIX_ACCESS_TOKEN}
# Only you can trigger it; and only with an explicit @hermes mention in shared
# rooms (DMs to it would respond unprompted, per Hermes's own default behavior).
MATRIX_ALLOWED_USERS: ${MATRIX_HUMAN_USER_ID}
MATRIX_REQUIRE_MENTION: "true"
# Routed through the local litellm gateway, not OpenRouter directly — same pattern
# as claude-agent's chat path, one place to hold the OpenRouter credential and swap
# models. Hermes's "main"/custom-endpoint provider is any OpenAI-compatible API
# reachable via OPENAI_BASE_URL + OPENAI_API_KEY. Note: this does NOT give Hermes
# access to the Claude Pro/Max subscription — that's blocked by Anthropic itself for
# any caller other than the real Claude Code CLI, proven earlier in this session
# (reproduced with plain curl straight to api.anthropic.com, LiteLLM or not).
OPENAI_BASE_URL: http://litellm:4000/v1
OPENAI_API_KEY: ${LITELLM_MASTER_KEY}
# Left disabled: Hermes itself warns that a network-reachable API server combined
# with the default unsandboxed ('local') terminal backend gives any caller full
# terminal/file access within the container. Matrix is the actual interface in use;
# re-enable (API_SERVER_HOST: 0.0.0.0) only alongside terminal.backend: docker if
# claude-agent ever needs to call Hermes programmatically.
API_SERVER_ENABLED: "false"
volumes:
- /home/william/hermes-data:/opt/data
networks:
- web
# Without this the image's default command launches the interactive CLI, which
# immediately exits ("Input is not a terminal") since a detached container has no
# stdin — the container then just sits there having done nothing, every restart.
command: ["gateway", "run"]
claude-agent:
image: ${GITEA_REGISTRY_IMAGE}
depends_on:
- litellm
container_name: claude-agent
restart: unless-stopped
# Explicit vars, not env_file: .env — Portainer's git-based stack deploy clones the
# repo fresh (no .env present, it's gitignored) and only performs ${VAR} substitution
# from the stack's own Env list, so env_file here would silently fail to load anything.
environment:
GITEA_URL: ${GITEA_URL}
GITEA_TOKEN: ${GITEA_TOKEN}
GITEA_WEBHOOK_SECRET: ${GITEA_WEBHOOK_SECRET}
# Claude subscription (Pro/Max) auth via `claude setup-token`, not API billing —
# Claude Code reads this in preference to ANTHROPIC_API_KEY when both could apply.
CLAUDE_CODE_OAUTH_TOKEN: ${CLAUDE_CODE_OAUTH_TOKEN}
MATRIX_HOMESERVER_URL: ${MATRIX_HOMESERVER_URL}
MATRIX_BOT_TOKEN: ${MATRIX_BOT_TOKEN}
MATRIX_CONTROL_ROOM_ID: ${MATRIX_CONTROL_ROOM_ID}
MATRIX_BOT_USER_ID: ${MATRIX_BOT_USER_ID}
OTHER_AGENT_USER_IDS: ${OTHER_AGENT_USER_IDS}
KNOWN_REPOS: ${KNOWN_REPOS}
# Chat replies go through the local litellm service (OpenRouter's models, incl. its
# auto-router), not OpenRouter directly.
LITELLM_BASE_URL: http://litellm:4000
LITELLM_MASTER_KEY: ${LITELLM_MASTER_KEY}
volumes:
- agent_workspace:/workspace
networks:
- web
labels:
- "traefik.enable=true"
- "traefik.http.routers.agent.rule=Host(`${AGENT_HOSTNAME}`)"
- "traefik.http.routers.agent.entrypoints=websecure"
- "traefik.http.routers.agent.tls.certresolver=letsencrypt"
- "traefik.http.services.agent.loadbalancer.server.port=3001"
act_runner:
image: gitea/act_runner:latest
container_name: act_runner
restart: unless-stopped
environment:
GITEA_INSTANCE_URL: ${GITEA_URL}
GITEA_RUNNER_REGISTRATION_TOKEN: ${ACT_RUNNER_REGISTRATION_TOKEN}
GITEA_RUNNER_NAME: gitops-vps-runner
# catthehacker/ubuntu:act-latest is the standard job-container image for
# act/act_runner — includes git + docker CLI, which job steps need.
GITEA_RUNNER_LABELS: docker:docker://catthehacker/ubuntu:act-latest
# No custom config.yaml: act_runner already auto-detects its own bind-mounted
# docker.sock and passes it through to job containers with no extra config needed.
# (A relative-path bind-mounted config file here would also be unreliable under
# Portainer's git-stack deploy — its cloned checkout doesn't persist for the
# container's runtime, so the mount source can silently resolve to nothing.)
volumes:
- /var/run/docker.sock:/var/run/docker.sock
- act_runner_data:/data
networks:
- web
networks:
web:
external: true
volumes:
matrix_data:
agent_workspace:
act_runner_data: