Flowcat feature-flag matrix

Every provider, transport, and exporter is a single Cargo feature, dep:-gated so the default build pulls none of their client dependencies. This file enumerates every feature across the four crates; it is derived from the Cargo.tomls and is the exhaustive companion to the README's summary table and PROVIDERS.md.

(D) = distinct client (own wire protocol). (W) = thin wrapper over a (D) family client (base_url + auth + default-model change). See CONTRIBUTING.md.

`flowcat-core` — `default = ["sip", "recorder"]`

Feature	Default	Pulls	What it gates
`sip`	✅	—	native SIP user-agent (REGISTER/INVITE/ACK/BYE)
`recorder`	✅	—	call recorder-as-processor (WAV via `hound`)
`vad-ort`	—	`ort`, `ndarray`	Silero VAD + Smart-Turn ONNX impls
`filter-rnnoise`	—	`nnnoiseless`	pure-Rust RNNoise noise-suppression filter

The trait seams (Transport/Stt/Tts/Llm/RealtimeLlm/Vad/Turn/FrameSerializer/ Brain/SessionSource), the Gemini Live client, codec/resample, and the SIP/RTP/SDP stack are always present; only the heavy ONNX/RNNoise bodies are gated.

`flowcat-services` — `default = []`

Nothing is on by default. Umbrellas: stt-all, tts-all, llm-all, realtime-all, obs-all.

Realtime / speech-to-speech (7 incl. core Gemini)

Feature	Tag	Pulls
`realtime-openai`	(D)	`tokio-tungstenite`, `tokio`, `base64`
`realtime-azure`	(W) over openai	—
`realtime-grok`	(W) over openai	—
`realtime-inworld`	(W) over openai	—
`realtime-ultravox`	(D)	`tokio-tungstenite`, `tokio`
`realtime-novasonic`	(D)	`tokio`, `base64`

Gemini Live is the 7th realtime impl and lives in flowcat-core (re-exported as flowcat_core::GeminiLive); it has no flowcat-services feature.

STT (20)

Feature	Tag	Transport
`stt-deepgram`	(D)	WS (ref impl)
`stt-assemblyai`	(D)	WS
`stt-gladia`	(D)	WS
`stt-soniox`	(D)	WS
`stt-speechmatics`	(D)	WS
`stt-cartesia`	(D)	WS
`stt-azure`	(D)	WS
`stt-gradium`	(D)	WS
`stt-elevenlabs`	(D)	segmented HTTP
`stt-sarvam`	(D)	REST
`stt-mistral`	(D)	REST
`stt-openai`	(D)	Whisper-HTTP base
`stt-groq` / `stt-fal` / `stt-speaches` / `stt-xai`	(W) over openai	Whisper-HTTP
`stt-google`	(D)	gRPC (`tonic`)
`stt-nvidia`	(D)	gRPC / Riva (`tonic`)
`stt-aws-transcribe`	(D)	SigV4 WS (`hmac`/`sha2`)
`stt-whisper-local`	(D)	local (`whisper-rs`; C build, needs `cmake`)

TTS (29)

Feature	Tag	Transport
`tts-cartesia`	(D)	WS (ref impl)
`tts-elevenlabs` / `tts-deepgram` / `tts-rime` / `tts-asyncai` / `tts-gradium` / `tts-soniox` / `tts-resemble`	(D)	WS
`tts-openai`	(D)	HTTP base
`tts-groq` / `tts-xai`	(W) over openai	OpenAI-TTS-HTTP
`tts-azure`	(D)	SSML WS/HTTP
`tts-sarvam` / `tts-mistral` / `tts-hume` / `tts-inworld` / `tts-minimax` / `tts-camb` / `tts-speechmatics`	(D)	HTTP
`tts-fish` / `tts-lmnt` / `tts-neuphonic` / `tts-smallest`	(D)	interruptible HTTP
`tts-kokoro` / `tts-piper` / `tts-xtts`	(D)	local/model HTTP
`tts-google` / `tts-nvidia`	(D)	gRPC (`tonic`)
`tts-aws-polly`	(D)	SigV4 (`hmac`/`sha2`)

LLM (23)

Feature	Tag
`llm-openai`	(D) — chat-completions SSE, ref impl
`llm-openai-responses`	(D) — Responses API
`llm-anthropic`	(D) — Messages API
`llm-google`	(D) — Gemini `generateContent`
`llm-aws-bedrock`	(D) — SigV4 event-stream (`hmac`/`sha2`)
`llm-groq`, `llm-together`, `llm-fireworks`, `llm-openrouter`, `llm-perplexity`, `llm-deepseek`, `llm-cerebras`, `llm-sambanova`, `llm-nebius`, `llm-novita`, `llm-qwen`, `llm-grok`, `llm-nvidia-nim`, `llm-ollama`, `llm-sarvam`, `llm-mistral`, `llm-azure`, `llm-speaches`	(W) over `llm-openai` (base_url + auth) — 18 wrappers

Observability + MCP

Feature	Pulls
`obs-otel`	`opentelemetry`
`obs-sentry`	`reqwest`
`obs-langfuse`	`reqwest`
`mcp`	`reqwest` — MCP-as-processor client

Brain adapters

Feature	Pulls
`brain-http`	`reqwest`, `tokio/rt-multi-thread` — `RemoteBrain`: drives conversation policy from an HTTP service (e.g. a Python webhook). See `examples/python-remote-brain`.

`flowcat-transports` — `default = []`

Feature	Pulls
`webrtc-str0m`	`str0m`, `audiopus`, `tokio`, `tokio-util`
`ws`	`tokio-tungstenite`, `tokio`, `tokio-util`
`daily`	`reqwest`
`livekit`	— (stub)
`local`	`audiopus`, `tokio` (local mic/speaker)

`flowcat-telephony` — `default = ["plivo"]`

Serializers are dependency-free flags (pure framing).

Feature	Default
`plivo`	✅
`twilio`, `telnyx`, `exotel`, `vonage`, `genesys`, `asterisk`, `cloudonix`, `vobiz`	—
`dtmf-inband`	— (in-band Goertzel DTMF; RFC2833 is always available)

`flowcat-agent` — `default = ["brain"]`

A declarative, graph-based agent: define an agent as a node/edge graph_spec (JSON/YAML) instead of hand-writing an AgentBrain. DeclarativeBrain is a ready-to-use flowcat_core::AgentBrain over the graph engine.

Feature	Pulls
`brain`	`flowcat-core` — the `DeclarativeBrain` `AgentBrain` adapter (default on)

With default-features = false the pure graph engine (parse / validate / {{var}} interpolation) builds with no flowcat-core dependency.

Toolchain caveats

gRPC (stt-google/stt-nvidia/tts-google/tts-nvidia) compiles .protos via tonic-build → needs protoc on PATH.
stt-whisper-local bundles whisper.cpp → needs cmake + a C/C++ toolchain. The Rust stub compiles; the dep's C build is the gate.

Everything else is rustls-only reqwest / tokio-tungstenite — no system OpenSSL, no AWS SDK (the AWS providers hand-roll SigV4 over hmac/sha2).