Advertising a capability and running a service are different claims, and conflating them made an Amethyst user un-addable to any group a White Noise user starts. The reference client installs agent-text-stream-quic-v1 with `required_member_roles = receive` into the required component set of EVERY group it creates, then refuses an invitee whose KeyPackage omits either the component or the `0xF2D1` role — checked before negotiation, so negotiation cannot rescue it, and the invite path applies the same rule. Dropping the advertisement on the grounds that nothing in the deployed network publishes previews was right about the traffic and wrong about the capability: a capability says "this client can handle it", never "this group uses it". So the component and the receive role go back. `send` and `fanout` stay off and no watcher is started — we can be shown a preview, we do not originate one, and nothing dials a broker. Two tests had encoded the old premise, one of them asserting the refusal as if it were a feature. They now assert the rule that actually decides interop: our default leaf is admitted by the reference's own stream policy, and is refused by a group that requires `send`. The KDoc on `currentProfileLeafCapabilities` had described the role as present the whole time — it was the code that had drifted from it. Co-Authored-By: Claude Opus 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_016kCuA6tc4JQzHPCDd39GHq
marmotQuic
Marmot's raw QUIC transport binding for agent text stream previews
(transports/quic.md), on top of the repo's own pure-Kotlin :quic stack.
Why this is not nestsClient's WebTransport
Both features move bytes over :quic, but they enter it at different layers.
nestsClient speaks WebTransport: HTTP/3, an Extended CONNECT handshake, a
:protocol pseudo-header, QPACK, SETTINGS negotiation. Its
WebTransportSession abstraction starts above all of that.
Marmot's binding is raw QUIC. It negotiates its own ALPN —
marmot.quic_broker.v1 for the broker path, marmot.quic_stream.v1 for the
direct one — and writes frames straight onto QUIC streams. There is no HTTP/3
in it at all, so WebTransportSession is the wrong shape.
What both share is everything below that line, which is the hard part and is already built: the QUIC connection, TLS 1.3, ALPN negotiation, stream multiplexing, loss recovery and the UDP socket.
Shape
- A publisher opens a client-initiated unidirectional stream, writes a
publishcontrol envelope, then record frames. - A subscriber opens a client-initiated bidirectional stream, writes a
subscribecontrol envelope, and reads the fan-out on the return direction. - A direct sender opens a client-initiated unidirectional stream and writes record frames with no control envelope — the dialed endpoint is already the one receiver, so there is no room to name.
A broker rejects the wrong pairing. Every role frames the same way:
uint32 frame_len || bytes — on the broker path the control envelope first and
then each AgentTextStreamRecordV1, on the direct path records from the first
byte.
Note the direct path's connection direction: the receiver listens and the
sender dials, inverted from the broker path where both ends dial the
broker. Only the sender half is here; :quic is a client stack with no server
role, so this module cannot expose a direct-path endpoint of its own.
The codecs — control envelope, frame reader/writer with both caps, quic://
candidate parsing — live in quartz next to the rest of agent-text-stream,
because they are pure bytes and belong with the feature. This module is only
the connection.
The broker sees nothing
Records are encrypted under a key derived from the group's MLS exporter. A
broker holds no key and learns only the routing pair
(stream_id, start_event_id) plus ciphertext. It is an untrusted forwarder,
and a candidate that points somewhere hostile still cannot forge a record.
TLS trust
Preview endpoints and brokers are commonly self-signed, and the binding says so:
a client MAY pin the endpoint certificate by exact DER or SHA-256 fingerprint
instead of chaining to a CA. PinnedCertificateValidator (in :quic) is that
pin. It replaces the chain and the hostname check and nothing else — the peer
still has to sign the TLS transcript with the pinned certificate's private key,
so copying a public certificate off the wire buys an attacker nothing.
amy marmot stream send|watch takes --pin-sha256 HEX[,HEX…]; the reference
broker prints its own server_cert_sha256_fingerprint in its startup JSON.
Interop tests
MarmotQuicBrokerInteropTest drives our publisher and subscriber through MDK's
own reference broker, and MarmotQuicDirectInteropTest drives our direct
sender against MDK's direct receiver (wn stream receive). Both are the only
way to know the binding is right: an ALPN string, a stream direction, a missing
control envelope and a frame prefix are all things an implementation will
happily agree with itself about.
Start the broker from an MDK checkout:
cargo build --release --bin marmot-quic-broker --bin wn
./target/release/marmot-quic-broker --bind 127.0.0.1:4450 --json
then:
./gradlew :marmotQuic:jvmTest \
-DmarmotQuicBroker=127.0.0.1:4450 \
-DmarmotQuicBrokerPin=<server_cert_sha256_fingerprint from that JSON> \
-DmarmotWn=/path/to/mdk/target/release/wn
Each property gates its own cases and they skip visibly without it, so an
ordinary ./gradlew test never needs the reference implementation on the
machine. -DmarmotWn needs no running process: the test spawns
wn stream receive itself on a free port.
Using it
amy marmot stream drives the whole feature; the harness's tests 18 and 19
run it in both directions against MDK.
amy marmot stream start GID --broker quic://127.0.0.1:4450
amy marmot stream send GID --stream-id … --start-event-id … --broker … "hello"
amy marmot stream send GID --stream-id … --start-event-id … --direct quic://host:port "hello"
amy marmot stream watch GID --stream-id …
amy marmot stream finish GID --stream-id … --transcript-hash … --chunk-count N "hello"
Advertised, but not started
The implementation is complete and tested, and nothing in the app starts it.
Those are two separate things, and conflating them broke interop once. Our
KeyPackage does advertise component 0x8006 and the 0xF2D1 receive role,
because the reference client installs agent-text-stream.quic.v1 with
required_member_roles = receive into the required set of every group it
creates, and refuses an invitee whose leaf omits either. Dropping the
advertisement on the grounds that "nothing publishes previews" made an Amethyst
user un-addable to any group a White Noise user started — the traffic claim was
right, the capability claim was not. A capability says "this client can handle
it", never "this group uses it".
What we do NOT advertise is send (0xF2D2) and fanout (0xF2D4): we can be
shown a preview, we do not originate one. A group that requires either refuses
us, and CurrentProfileWelcomeTest asserts that refusal so widening the default
stays a deliberate decision.
What is not started: the Android chat screen builds no watcher and dials no
broker a kind:1200 advertises. The codecs, this module, the CLI
(amy marmot stream …) and the interop tests all still work and still run.
Turning the live path on is re-adding the watcher to MarmotGroupChatView.
Not done
- The Android GUI renders previews but does not originate a stream — that is
an agent's job, and no agent runs in the app yet. Only
amypublishes one. - The desktop app has no Marmot chat screen at all, so there is nothing to
render a preview into. The watcher it would use already lives in
commons. - The direct path's receiving half.
:quichas no server role, so this module can dial a direct receiver but cannot be one. v1 also defines no start-payload candidate format for the direct path, so a sender only reaches a receiver whose endpoint it already knows out of band.