Atlas
m4/org/openalice-voiceci failure

openalice-voice

[Atlas card](https://atlas.blal.pro/repos/openalice-voice) · Live at **`https://voice.blal.pro`**

Features
9
Status
·unknown
Last activity
1mo ago
Branch
mvp

Architecture rating

★★★★☆C+1mo ago
consistency★★★★☆★B

Manifest valid (100/100), zero drift, no circular deps, 0 dead imports — Atlas quality 91/A shows strong substrate/convention adoption across signal/rtp/tts/stt module layout.

testing★★★★☆C+

26 integration tests thoroughly cover the critical WS signalling path (join/leave/mute/relay bi-dir/ICE, multi-client broadcast, disconnect cleanup) plus rooms/tokens/metrics — but CI run is failing and no unit tests or TTS/STT-provider/SSRF regression coverage is visible.

observability★★★★☆C+

/health probe live (200) and Prometheus-style /metrics exposing voice_rooms_active gauge (tested), but no distributed-tracing or confirmed structured-logging evidence for the WS/RTP paths.

security★★★★☆C+

Internal-secret gating on /api/v1/say + GDPR export, room JWT issuance via /api/v1/tokens, DTLS/SRTP media path — but no evidence of rate-limiting on TTS/room creation or security headers, and WebRTC signalling SSRF/abuse guards unverified.

docs★★★★☆C+

Strong manifest (100) with 9 @feature annotations (92) plus AGENTS.md, docs/ dir and README/Atlas card — undercut by 236 missing docstrings driving code_health to 50/100 (the repo's weakest axis).

data layer★★★☆★☆C

Stateless SFU owns zero tables (by design) so no schema/migrations/RLS to grade; GDPR TTS-cache export endpoint (/internal/me/:tenant_id/export.json) is present and tenant-scoped, but cache persistence/retention controls are unverified.

reliability★★★☆★☆C

Health probe 200 with 100% 24h uptime and tested WS disconnect/participant cleanup, but CI is currently RED (run failed 2026-06-16), no graceful-SIGTERM-shutdown evidence, and 16 complexity flags + 18 god-modules raise fault-surface concerns.

source · auto-grader-2026-06-16

Ecosystem manifest

.atlas-deps.yml

Rust SFU media server for real-time voice rooms (LiveKit-inspired): WebRTC/ICE/DTLS/SRTP signalling, pluggable TTS/STT providers, and transcript dispatch to live-control and social-api.

Runtime deps
  • openalice-agent-protocol-rsother
    via cargo
    AgentRuntimeConfig + fetch_agent_config() helper for deserialising the unified agent config returned by openalice-auth's internal endpoint.
  • openalice-authhttp
    via AUTH_INTERNAL_URL
    Fetch per-user unified agent_config (TTS/STT provider, voice id, emotion, pace, language) via /internal/agent-config/:user_id.
  • openalice-axum-commonother
    via cargo
    Shared axum plumbing: internal_client() builder and verify_internal_secret() used by every inbound /api/v1/say and /internal/* call.
  • openalice-live-controlhttpoptional
    via VOICE_DISPATCH_URL
    POST /internal/agent/dispatch with ConnectorEvent::VoiceUtterance once the SFU's STT audio-capture path is wired (TranscriptDispatcher).
  • openalice-social-apihttpoptional
    via VOICE_DISPATCH_URL
    Same dispatch endpoint — product=social path. Optional: voice does not need social-api at startup.
Consumed env
  • VOICE_JWT_SECRET
  • SHARED_JWT_SECRET
  • VOICE_ALLOW_INSECURE
  • INTERNAL_BROADCAST_SECRET
  • INTERNAL_SECRET
  • VOICE_API_KEY
  • VOICE_API_SECRET
  • VOICE_PORT
  • VOICE_UDP_PORT
  • VOICE_BIND
  • VOICE_CORS_ORIGINS
  • VOICE_RATE_LIMIT
  • VOICE_MAX_ROOMS
  • VOICE_MAX_PARTICIPANTS
  • TTS_PROVIDER
  • ELEVENLABS_API_KEY
  • ELEVENLABS_BASE_URL
  • ELEVENLABS_VOICE_ID
  • ELEVENLABS_MODEL_ID
  • ELEVENLABS_STREAMING_MODEL_ID
  • CARTESIA_API_KEY
  • CARTESIA_BASE_URL
  • CARTESIA_MODEL
  • CARTESIA_DEFAULT_VOICE
  • CARTESIA_VOICE_ID
  • OPENAI_API_KEY
  • OPENAI_BASE_URL
  • OPENAI_REALTIME_API_KEY
  • OPENAI_REALTIME_BASE_URL
  • OPENAI_REALTIME_MODEL
  • OPENAI_REALTIME_VOICE
  • OPENAI_TTS_MODEL
  • OPENAI_TTS_VOICE
  • HUME_API_KEY
  • HUME_BASE_URL
  • HUME_DEFAULT_VOICE
  • GOOGLE_GEMINI_API_KEY
  • GEMINI_API_KEY
  • GEMINI_BASE_URL
  • GEMINI_TTS_MODEL
  • GEMINI_DEFAULT_VOICE
  • OPENROUTER_API_KEY
  • OPENROUTER_TTS_MODEL_ID
  • DASHSCOPE_API_KEY
  • DASHSCOPE_BASE_URL
  • QWEN3_API_KEY
  • QWEN3_TTS_MODEL
  • QWEN3_DEFAULT_VOICE
  • SESAME_ENDPOINT
  • STT_PROVIDER
  • DEEPGRAM_API_KEY
  • DEEPGRAM_MODEL
  • WHISPER_MODEL
  • ASSEMBLYAI_API_KEY
  • VOICE_CACHE_DIR
  • OTEL_EXPORTER_OTLP_ENDPOINT
  • OTEL_SERVICE_NAME
Exposes
  • docker_image: openalice-voice
  • http_endpoint: POST /api/v1/say — TTS synthesis (internal-secret-gated)
  • http_endpoint: POST /api/v1/say-stream — streaming TTS + lipsync SSE (ElevenLabs)
  • http_endpoint: POST /api/v1/rooms — create voice room
  • http_endpoint: GET /api/v1/rooms — list voice rooms
  • http_endpoint: POST /api/v1/tokens — issue room JWT
  • ws_endpoint: WS /ws — WebRTC signalling (join/leave/offer/answer/ICE)
  • http_endpoint: GET /internal/me/:tenant_id/export.json — GDPR TTS-cache export (internal-secret-gated)

Roadmap · what's planned next

0% avg5 goals
  • alice-integration-default
    planned
    0%
    no description
  • clustering-redis-default
    planned
    0%
    no description
  • stt-transcript-dispatch
    planned
    0%
    no description
  • webrtc-rs-migration
    planned
    0%
    no description
  • per-tenant-tts-quota
    planned
    0%
    no description

Code composition

tokei · loc
7.3kloc
  • Rust99.1%
  • TOML0.9%
  • Rust99.1%· 7,204
  • TOML0.9%· 68
Language% codecodecommentsblanksfiles
Rust99.1%7,2046671,04745
TOML0.9%6850212
Markdown0.0%03361044

Telemetry ribbon · last 90 days

10 events
MAYJUNJUL2026-05-21 · commit · new commit ed55574 → aaaf45a2026-05-21 · commit · new commit aaaf45a → 358b2dd2026-05-21 · commit · new commit 358b2dd → 95aabc22026-05-21 · commit · new commit 95aabc2 → 2b94a1c2026-05-22 · commit · new commit 2b94a1c → ef683d92026-05-22 · commit · new commit ef683d9 → cc20cb32026-05-23 · commit · new commit cc20cb3 → dde9cc52026-06-16 · commit · new commit dde9cc5 → 1e5d8bc2026-06-16 · commit · new commit 1e5d8bc → 95d55e62026-06-16 · commit · new commit 95d55e6 → a016984NOW
commitmanifestlifecyclekind

voice/agents

1 feature
  • voice.alice.tools
    src/alice.rs:61
    Voice tool schemas (voice_join, voice_leave, voice_speak, voice_status) for AI agents to participate in rooms.
    since 0.1.0

voice/audio

1 feature
  • voice.speaker.detect
    src/rtp/speaker.rs:58
    Active speaker detection using LiveKit 2-gate + EMA over RFC 6464 audio levels.
    since 0.1.0

voice/auth

1 feature
  • voice.auth.token
    src/lib.rs:138
    Issue a signed JWT (HS256) for joining a room with publish/subscribe scopes.
    api:POST /api/v1/tokens
    since 0.1.0

voice/cluster

1 feature
  • voice.cluster.select-node
    src/cluster.rs:77
    Pick the best node for a new room using strategy (Any, LowestSysload, TwoChoice, RegionAware).
    since 0.1.0

voice/quality

1 feature
  • voice.quality.mos
    src/rtp/quality.rs:31
    Per-participant connection quality MOS score using LiveKit E-model with asymmetric smoothing.
    since 0.1.0

voice/rooms

2 features
  • voice.rooms.create
    src/lib.rs:73
    Create a new voice room with a name and participant cap.
    api:POST /api/v1/rooms
    since 0.1.0
  • voice.rooms.list
    src/lib.rs:103
    List active voice rooms with participant counts.
    api:GET /api/v1/rooms
    since 0.1.0

voice/signal

1 feature
  • voice.signal.websocket
    src/lib.rs:209
    WebSocket signalling channel for join/leave/mute/offer/answer/ICE.
    api:WS /ws
    since 0.1.0

voice/transport

1 feature
  • voice.transport.udp-mux
    src/transport/mod.rs:91
    Single-port UDP mux with RFC 5764 packet classification (STUN/DTLS/RTP/RTCP).
    since 0.1.0