[FEATURE] Security Phase 2 - PyTorch 2.6, Docker hardening, env migration by POWERFULMOVES · Pull Request #443 · POWERFULMOVES/PMOVES.AI

POWERFULMOVES · 2026-01-03T16:53:55Z

Summary

Rebased PR with 13 commits from `feat/hardened-security-phase2` not yet in hardened.

Changes

Security Hardening

Upgrade PyTorch to 2.6.0 (Dependabot Install libsndfile dependencies in audio processor image #61) (fix(security): Upgrade PyTorch to 2.6.0 (Dependabot #61) #344)
Docker hardening for notebook-sync and vibevoice-realtime
GPU access controls and production hardening (fix(security): GPU access + production hardening (health checks, image pins, required password) #355)
Health checks, image pins, required password enforcement

Environment Migration (6-Tier Architecture)

Phase 2: Migrate API tier to env-tier-api anchor (refactor(env): Phase 2 - Migrate API tier to env-tier-api anchor #350)
Phase 3: Migrate LLM tier to env-tier-llm anchor (refactor(env): Phase 3 - Migrate LLM tier to env-tier-llm anchor #351)
Phase 6: Migrate agent tier + API additions (refactor(env): Phase 6 - Migrate agent tier + complete tier consolidation #354)

Infrastructure Fixes

Agent Zero settings persistence and Archon hostname
Archon Supabase connectivity and UI dashboard fixes
Gateway Agent silent exception logging (fix(flute-gateway): add logging to silent exception handlers #323)
DeepResearch Prometheus counter double-registration fix
WSL2 playback hardening and Makefile voice-say target (fix(voice): harden WSL2 playback and Makefile voice-say target #320)

Documentation

Comprehensive service documentation (docs: Add comprehensive service documentation #338)

Testing

```bash
docker compose -p pmoves config --quiet
curl -sf http://localhost:8080/healthz # Agent Zero
curl -sf http://localhost:8091/healthz # Archon
```

Phase 2 of env consolidation: API tier migration Changes: - Fix tier anchor syntax (block-style env_file with required: false) - Migrate postgrest, presign, retrieval-eval to <<: *env-tier-api - Create env.tier-api.example with secure defaults API tier services receive data tier URLs and internal credentials only, no external API keys (OPENAI_API_KEY, etc.) Services migrated: - postgrest: PostgREST database gateway - presign: MinIO URL presigner - retrieval-eval: Hi-RAG evaluation service 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Phase 3 of env consolidation: LLM tier migration Changes: - Fix tier anchor syntax (block-style env_file with required: false) - Migrate tensorzero-ui to <<: *env-tier-llm - Create env.tier-llm.example with all LLM provider API keys CRITICAL: LLM tier is the ONLY tier with access to external API keys. All other services call TensorZero internally, not providers directly. Services in LLM tier: - tensorzero-gateway: Already using <<: *env-tier-llm - tensorzero-ui: Migrated from legacy env_file - tensorzero-clickhouse: Uses inline env (no API keys needed) - pmoves-ollama: Uses inline env (no API keys needed) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Completes the 6-tier environment consolidation by migrating: **Agent Tier (10 services):** - mesh-agent, deepresearch, supaserch, consciousness-service - archon-agent-work-orders, botz-gateway, publisher-discord - messaging-gateway, jellyfin-bridge, chat-relay **API Tier Additions (4 services):** - postgrest-cli, hi-rag-gateway, hi-rag-gateway-gpu, hi-rag-gateway-v2-gpu **Worker Tier (6 services) - included in this commit:** - render-webhook, comfy-watcher, pdf-ingest, langextract - notebook-sync, session-context-worker **Media Tier (10 services) - included in this commit:** - ultimate-tts-studio, flute-gateway, ffmpeg-whisper - media-video, media-audio, channel-monitor - invidious, invidious-companion, grayjay-plugin-host, grayjay-server Security improvement: 30 services now use tier-based env_file anchors instead of legacy x-env-legacy pattern. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

…e pins, required password) (#355) * docs: address PR review comments for env tier consolidation - Add GPU Orchestrator and E2B Runner to services-catalog.md - Add gpu-orchestrator to env.tier-api.example service list - Create learnings file documenting 6-tier env architecture - Add env.tier-*.example files for worker, media, agent tiers Addresses nitpick comments from PRs #349-354. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(gpu): ensure GPU access for all CUDA-enabled services - gpu-orchestrator: Add GPU deploy section + NVIDIA_VISIBLE_DEVICES - ultimate-tts-studio: Add NVIDIA_VISIBLE_DEVICES env var - hi-rag-gateway-gpu: Add NVIDIA_VISIBLE_DEVICES env var - hi-rag-gateway-v2-gpu: Add NVIDIA_VISIBLE_DEVICES env var - media-audio: Change base image to nvidia/cuda:12.4.1-runtime-ubuntu22.04 (was python:3.11-slim which caused silent CPU fallback) This fixes silent CPU fallbacks where PyTorch CUDA packages were installed but the CUDA runtime was not available in the container. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(security): production hardening - health checks and image pinning Critical security and reliability fixes: 1. **POSTGRES_PASSWORD now required** - Changed from insecure default `:-pmoves` to required syntax `${VAR:?error}`. Compose will fail if POSTGRES_PASSWORD is not set (lines 59, 78, 937). 2. **Pinned 8 :latest images to specific versions:** - postgrest/postgrest:latest → v12.2.3 - minio/minio:latest → RELEASE.2024-12-18T13-15-44Z - ollama/ollama:latest → 0.5.4 - tensorzero/gateway:latest → 2024.12.18 - tensorzero/ui:latest → 2024.12.18 - invidious:latest → 2024.12.09 - invidious-companion:latest → 2024.12.20 - grayjay:latest → 2024.11.01 3. **Added health checks to 37 services** (52 total, up from 15): - Data tier: qdrant, meilisearch, minio - API tier: hi-rag-*, retrieval-eval, presign, render-webhook - Worker tier: extract-worker, pdf-ingest, langextract, notebook-sync, ffmpeg-whisper, media-video, media-audio, pmoves-yt, channel-monitor - Agent tier: agent-zero, mesh-agent, deepresearch, supaserch - TensorZero: gateway, ui, ollama - Others: publisher-discord, messaging-gateway, jellyfin-bridge 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(health): address PR review critical issues Fixes critical issues identified by PR review: MinIO: - Change health check from `mc ready local` to curl-based (minio/minio image doesn't include mc binary) NATS: - Add `-m 8222` flag to enable HTTP monitoring port (health check was targeting port that wasn't enabled) Health checks: - comfy-watcher: verify module imports instead of just `import sys` - mesh-agent: verify main module and NATS client availability Missing start_period: - postgres: add 15s start_period - chat-relay: add 30s start_period - n8n-agent: add 30s start_period - invidious-postgres: add 15s start_period All 52 health checks now have start_period defined. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> --------- Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Increase Dockerfile hardening from 33/36 (91.7%) to 35/36 (97.2%): - notebook-sync: Add USER pmoves (user creation already existed) - vibevoice-realtime: Add full hardening with UID/GID 65532 Accepted exception: agent-zero uses root for initialization then drops to pmoves via 'su' for the service process. This pattern is required by upstream Agent Zero's prepare.py and /ins/copy_A0.sh. Pattern reference: flute-gateway/Dockerfile 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

- Normalize Supabase Kong hostnames to lowercase (DNS convention) - Remove duplicate ARCHON_SUPABASE_BASE_URL env var definition - Fix notebook-sync Dockerfile: remove USER directive so entrypoint can run chown/su as root before dropping privileges - Improve deepresearch _get_or_create_counter: use module-level cache to avoid private prometheus_client API dependency 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Changes: - Archon health check: Add Kong gateway hostname support for internal Docker networking (supabase_kong_pmoves.ai:8000) - UI dashboard: Add /dashboard redirect page to /dashboard/services - Grafana: Fix env var placeholders and job name mismatches in queries - Docs: Add placeholder for Jellyfin service documentation The Archon health check now properly handles both: - Supabase CLI endpoint (host.docker.internal:65421) - Internal Kong gateway (supabase_kong_pmoves.ai:8000) This resolves the 404 errors in Archon health checks when using internal Docker DNS names instead of host.docker.internal. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Add pmoves/data/agent-zero/ to gitignore. This directory contains runtime settings (settings.json) that persist via Docker volume mount. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

Use registry lookup pattern instead of try/except for counter registration. Prevents 'Duplicated timeseries' error when container restarts with existing registry state. Before: try/except around Counter() creation After: REGISTRY._names_to_collectors.get() for existing lookup 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

- Add warning log to Supabase health check exception handler - Improve persona fetch error logging with status code and truncated body - Add metrics tracking for non-200 persona fetch responses Closes #322 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

Merged after rebase to resolve Jellyfin README conflict

- Add volume mount ./data/agent-zero/tmp:/a0/tmp for settings.json Root cause: PMOVES-Agent-Zero/python/helpers/settings.py:162 stores settings at /a0/tmp/settings.json but path was not mounted - Normalize Archon SUPABASE_URL to lowercase supabase_kong_pmoves.ai Root cause: Docker DNS is case-sensitive on some configurations - Fix archon-agent-work-orders health check: /healthz → /health Root cause: Service exposes /health, not /healthz 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

chatgpt-codex-connector · 2026-01-03T16:54:00Z

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

coderabbitai · 2026-01-03T16:54:01Z

Important

Review skipped

Auto reviews are disabled on base/target branches other than the default branch.

Please check the settings in the CodeRabbit UI or the .coderabbit.yaml file in this repository. To trigger a single review, invoke the @coderabbitai review command.

You can disable this status message by setting the reviews.review_status to false in the CodeRabbit configuration file.

Thanks for using CodeRabbit! It's free for OSS, and your support helps us grow. If you like it, consider giving us a shout-out.

❤️ Share

_{Comment @coderabbitai help to get the list of available commands and usage tips.}

…tion (#443) * refactor(env): Phase 2 - Migrate API tier to env-tier-api anchor (#350) Phase 2 of env consolidation: API tier migration Changes: - Fix tier anchor syntax (block-style env_file with required: false) - Migrate postgrest, presign, retrieval-eval to <<: *env-tier-api - Create env.tier-api.example with secure defaults API tier services receive data tier URLs and internal credentials only, no external API keys (OPENAI_API_KEY, etc.) Services migrated: - postgrest: PostgREST database gateway - presign: MinIO URL presigner - retrieval-eval: Hi-RAG evaluation service 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * refactor(env): Phase 3 - Migrate LLM tier to env-tier-llm anchor (#351) Phase 3 of env consolidation: LLM tier migration Changes: - Fix tier anchor syntax (block-style env_file with required: false) - Migrate tensorzero-ui to <<: *env-tier-llm - Create env.tier-llm.example with all LLM provider API keys CRITICAL: LLM tier is the ONLY tier with access to external API keys. All other services call TensorZero internally, not providers directly. Services in LLM tier: - tensorzero-gateway: Already using <<: *env-tier-llm - tensorzero-ui: Migrated from legacy env_file - tensorzero-clickhouse: Uses inline env (no API keys needed) - pmoves-ollama: Uses inline env (no API keys needed) 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * refactor(env): Phase 6 - Migrate agent tier + API additions (#354) Completes the 6-tier environment consolidation by migrating: **Agent Tier (10 services):** - mesh-agent, deepresearch, supaserch, consciousness-service - archon-agent-work-orders, botz-gateway, publisher-discord - messaging-gateway, jellyfin-bridge, chat-relay **API Tier Additions (4 services):** - postgrest-cli, hi-rag-gateway, hi-rag-gateway-gpu, hi-rag-gateway-v2-gpu **Worker Tier (6 services) - included in this commit:** - render-webhook, comfy-watcher, pdf-ingest, langextract - notebook-sync, session-context-worker **Media Tier (10 services) - included in this commit:** - ultimate-tts-studio, flute-gateway, ffmpeg-whisper - media-video, media-audio, channel-monitor - invidious, invidious-companion, grayjay-plugin-host, grayjay-server Security improvement: 30 services now use tier-based env_file anchors instead of legacy x-env-legacy pattern. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * fix(security): GPU access + production hardening (health checks, image pins, required password) (#355) * docs: address PR review comments for env tier consolidation - Add GPU Orchestrator and E2B Runner to services-catalog.md - Add gpu-orchestrator to env.tier-api.example service list - Create learnings file documenting 6-tier env architecture - Add env.tier-*.example files for worker, media, agent tiers Addresses nitpick comments from PRs #349-354. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(gpu): ensure GPU access for all CUDA-enabled services - gpu-orchestrator: Add GPU deploy section + NVIDIA_VISIBLE_DEVICES - ultimate-tts-studio: Add NVIDIA_VISIBLE_DEVICES env var - hi-rag-gateway-gpu: Add NVIDIA_VISIBLE_DEVICES env var - hi-rag-gateway-v2-gpu: Add NVIDIA_VISIBLE_DEVICES env var - media-audio: Change base image to nvidia/cuda:12.4.1-runtime-ubuntu22.04 (was python:3.11-slim which caused silent CPU fallback) This fixes silent CPU fallbacks where PyTorch CUDA packages were installed but the CUDA runtime was not available in the container. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(security): production hardening - health checks and image pinning Critical security and reliability fixes: 1. **POSTGRES_PASSWORD now required** - Changed from insecure default `:-pmoves` to required syntax `${VAR:?error}`. Compose will fail if POSTGRES_PASSWORD is not set (lines 59, 78, 937). 2. **Pinned 8 :latest images to specific versions:** - postgrest/postgrest:latest → v12.2.3 - minio/minio:latest → RELEASE.2024-12-18T13-15-44Z - ollama/ollama:latest → 0.5.4 - tensorzero/gateway:latest → 2024.12.18 - tensorzero/ui:latest → 2024.12.18 - invidious:latest → 2024.12.09 - invidious-companion:latest → 2024.12.20 - grayjay:latest → 2024.11.01 3. **Added health checks to 37 services** (52 total, up from 15): - Data tier: qdrant, meilisearch, minio - API tier: hi-rag-*, retrieval-eval, presign, render-webhook - Worker tier: extract-worker, pdf-ingest, langextract, notebook-sync, ffmpeg-whisper, media-video, media-audio, pmoves-yt, channel-monitor - Agent tier: agent-zero, mesh-agent, deepresearch, supaserch - TensorZero: gateway, ui, ollama - Others: publisher-discord, messaging-gateway, jellyfin-bridge 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(health): address PR review critical issues Fixes critical issues identified by PR review: MinIO: - Change health check from `mc ready local` to curl-based (minio/minio image doesn't include mc binary) NATS: - Add `-m 8222` flag to enable HTTP monitoring port (health check was targeting port that wasn't enabled) Health checks: - comfy-watcher: verify module imports instead of just `import sys` - mesh-agent: verify main module and NATS client availability Missing start_period: - postgres: add 15s start_period - chat-relay: add 30s start_period - n8n-agent: add 30s start_period - invidious-postgres: add 15s start_period All 52 health checks now have start_period defined. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> --------- Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * security(dockerfile): Harden notebook-sync and vibevoice-realtime Increase Dockerfile hardening from 33/36 (91.7%) to 35/36 (97.2%): - notebook-sync: Add USER pmoves (user creation already existed) - vibevoice-realtime: Add full hardening with UID/GID 65532 Accepted exception: agent-zero uses root for initialization then drops to pmoves via 'su' for the service process. This pattern is required by upstream Agent Zero's prepare.py and /ins/copy_A0.sh. Pattern reference: flute-gateway/Dockerfile 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(infra): Address PR #345 CodeRabbit review comments - Normalize Supabase Kong hostnames to lowercase (DNS convention) - Remove duplicate ARCHON_SUPABASE_BASE_URL env var definition - Fix notebook-sync Dockerfile: remove USER directive so entrypoint can run chown/su as root before dropping privileges - Improve deepresearch _get_or_create_counter: use module-level cache to avoid private prometheus_client API dependency 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(infra): Archon Supabase connectivity and UI dashboard fixes Changes: - Archon health check: Add Kong gateway hostname support for internal Docker networking (supabase_kong_pmoves.ai:8000) - UI dashboard: Add /dashboard redirect page to /dashboard/services - Grafana: Fix env var placeholders and job name mismatches in queries - Docs: Add placeholder for Jellyfin service documentation The Archon health check now properly handles both: - Supabase CLI endpoint (host.docker.internal:65421) - Internal Kong gateway (supabase_kong_pmoves.ai:8000) This resolves the 404 errors in Archon health checks when using internal Docker DNS names instead of host.docker.internal. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * chore(gitignore): Exclude agent-zero runtime data directory Add pmoves/data/agent-zero/ to gitignore. This directory contains runtime settings (settings.json) that persist via Docker volume mount. 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(deepresearch): Prometheus counter double-registration Use registry lookup pattern instead of try/except for counter registration. Prevents 'Duplicated timeseries' error when container restarts with existing registry state. Before: try/except around Counter() creation After: REGISTRY._names_to_collectors.get() for existing lookup 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> * fix(flute-gateway): add logging to silent exception handlers (#323) - Add warning log to Supabase health check exception handler - Improve persona fetch error logging with status code and truncated body - Add metrics tracking for non-200 persona fetch responses Closes #322 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com> * docs: add comprehensive service documentation (#338) Merged after rebase to resolve Jellyfin README conflict * fix(compose): Agent Zero settings persistence and Archon hostname - Add volume mount ./data/agent-zero/tmp:/a0/tmp for settings.json Root cause: PMOVES-Agent-Zero/python/helpers/settings.py:162 stores settings at /a0/tmp/settings.json but path was not mounted - Normalize Archon SUPABASE_URL to lowercase supabase_kong_pmoves.ai Root cause: Docker DNS is case-sensitive on some configurations - Fix archon-agent-work-orders health check: /healthz → /health Root cause: Service exposes /health, not /healthz 🤖 Generated with [Claude Code](https://claude.com/claude-code) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com> --------- Co-authored-by: Codex Agent <codex-agent@example.com> Co-authored-by: Claude Opus 4.5 <noreply@anthropic.com>

- Update error modal to show default 'docker compose up --build -d' command - Add better organized note structure with bullet points - Include profile-specific fallback example for existing users - Update README Quick Start to show default command first - Maintain backward compatibility guidance for profile users

POWERFULMOVES and others added 12 commits January 3, 2026 11:52

docs: add comprehensive service documentation (#338)

972ccca

Merged after rebase to resolve Jellyfin README conflict

POWERFULMOVES merged commit 572c073 into PMOVES.AI-Edition-Hardened Jan 3, 2026
2 of 3 checks passed

POWERFULMOVES deleted the feat/hardened-security-phase2 branch March 7, 2026 21:44

Navigation Menu

[FEATURE] Security Phase 2 - PyTorch 2.6, Docker hardening, env migration - #443

[FEATURE] Security Phase 2 - PyTorch 2.6, Docker hardening, env migration#443
POWERFULMOVES merged 12 commits into
PMOVES.AI-Edition-Hardenedfrom
feat/hardened-security-phase2

POWERFULMOVES commented Jan 3, 2026

Uh oh!

chatgpt-codex-connector Bot commented Jan 3, 2026

Uh oh!

coderabbitai Bot commented Jan 3, 2026

Review skipped

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant

Conversation

POWERFULMOVES commented Jan 3, 2026

Summary

Changes

Security Hardening

Environment Migration (6-Tier Architecture)

Infrastructure Fixes

Documentation

Testing

Related

Uh oh!

chatgpt-codex-connector Bot commented Jan 3, 2026

Uh oh!

coderabbitai Bot commented Jan 3, 2026

Review skipped

Uh oh!

Uh oh!

Reviewers

Assignees

Labels

Projects

Milestone

Development

Uh oh!

1 participant