Skip to content

feat(binding-llm): server framing decode and forward - #2569

Open
jfallows wants to merge 18 commits into
developfrom
claude/stoic-sagan-9d81tf
Open

feat(binding-llm): server framing decode and forward#2569
jfallows wants to merge 18 commits into
developfrom
claude/stoic-sagan-9d81tf

Conversation

@jfallows

@jfallows jfallows commented Sep 11, 2026

Copy link
Copy Markdown
Contributor

Description

LlmServerFactory — the llm server stream factory. On BEGIN, resolves the inbound dialect via LlmDialectResolver (from the :path/headers of the accepted HttpBeginExFW), selects the matching LlmContentDecoder from LlmContentDecoderFactory keyed by the dialect's content-type, and forwards decoded canonical DATA/FLUSH frames to the routed application binding, with LlmBeginEx.dialect annotated on the forwarded BEGIN.

  • Framing decode runs unconditionally on every route, same-dialect included, per the design's layering rule — no dialect translation happens here (that's llm client's job when dialects differ, tracked separately).
  • When no decoder is registered for the resolved content-type (e.g. non-streaming JSON, for which no LlmContentDecoderSpi implementation exists yet), the body is forwarded opaquely as DATA followed by a terminal raw FLUSH on END, rather than guessing at an undefined decode contract.
  • The reply direction (application → network) is a generic, unmodified byte passthrough — no re-encoding logic is implemented here, since that isn't this issue's scope either.
  • An unroutable or unresolved-dialect request is rejected by returning null from newStream, relying on the engine's existing auto-reset behavior rather than a bespoke reject handler.
  • Modelled on binding-mcp's server-side stream handling patterns (buffer-slot decode/resume, *State bitmask, generic do* frame builders).

Covered by new k3po specs (sse.passthrough, opaque.fallback) under streams/network/streams/application, each with paired runtime IT (LlmServerIT, against a live engine) and peer-to-peer IT (NetworkIT/ApplicationIT, scripts run directly against each other) coverage, plus a unit test for the new k3po EL helper functions (LlmFunctionsTest).

Fixes #2484

This branch stacks on #2568 (dialect detection dispatcher and fixed-dialect config, #2483), which itself stacks on the earlier M1 protocol-scaffold PRs — so this diff includes those commits until they merge to develop, at which point this PR's diff will shrink to just this framing-decode commit.

Generated by Claude Code

@claude
feat(binding-llm.spec): define llm.idl stream extension types
cb31300 
Scaffold incubator/binding-llm.spec per AGENTS.md conventions and define
LlmBeginEx, LlmDataEx, and the LlmFlushEx union, modelled on
binding-mcp.spec's idl.

LlmBeginEx carries dialect only; model routing is deferred. LlmDataEx has
no fields: content flows through the DATA frame's own payload octets and
INIT/FIN through its existing flags, so nothing survives in the extension
once block identity moves to the FLUSH plane. LlmFlushEx is a 7-case union
covering message start, block start/end, finish, usage, keepalive, and an
opaque native/raw case for re-encoding events a same-dialect route doesn't
recognize.

Fixes #2476

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AuZoMsETwEczJx3cbkb8EJ
@claude
feat(binding-llm): module scaffold with Incubating LlmBindingInfo
daf8a87 
Scaffolds incubator/binding-llm and incubator/binding-llm.conf, modelled
on binding-mcp's SERVER/CLIENT BindingContext structure. LlmBindingInfo
is annotated @Incubating so type: llm config loading is gated behind
ZILLA_INCUBATOR_ENABLED via FeatureFilter, matching the AmqpBindingInfo/
PgsqlBindingInfo/RisingwaveBindingInfo precedent.

Fixes #2477

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_0142buJWS7C89AKr9uDJtSy4
@claude
feat(binding-llm): LlmContentDecoderSpi internal SPI, keyed by conten…
b822228 
…t-type

Registers by content-type and hands back a per-stream LlmContentDecoder;
stays in an internal, unexported package for now with no concrete
implementation registered yet.

Fixes #2478

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Mw32oxEw24fLt5Ypakj6pH
@jfallows jfallows mentioned this pull request Sep 11, 2026
Open
@claude
feat(binding-llm): application/json content decoder
8faad28 
Closes the non-streaming half of #2478's own scope: "Non-streaming
application/json goes through the same abstraction as one event, rather
than a special-cased branch." Only text/event-stream had an
LlmContentDecoderSpi implementation; application/json requests
(non-streaming dialect responses) had no decoder to dispatch to.

LlmJsonContentDecoder treats the entire buffered document as a single
event (one data + one flush call, no framing loop), mirroring
LlmSseContentDecoder's structure and unit-test conventions.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Mw32oxEw24fLt5Ypakj6pH
@jfallows
jfallows force-pushed the claude/stoic-sagan-9d81tf branch 2 times, most recently from 0510c5d to 5f11258 Compare September 12, 2026 04:33
@claude
refactor(binding-llm): collapse decoder/encoder SPIs into LlmContentC…
0d82f5d 
…odecSpi

LlmContentDecoderSpi and LlmContentEncoderSpi (the latter previously
living downstream in #2571) let a content-type register a decoder with
no matching encoder, or vice versa, since each was ServiceLoader-discovered
and dispatched independently. LlmContentCodecSpi makes "this content-type
is fully supported" one type-enforced fact: a single contentType() key
with both supplyDecoder() and supplyEncoder(), one META-INF/services
registration per content-type, one LlmContentCodecFactory dispatching
both directions.

Pulls the internal/encode/ base package and its text/event-stream
implementation (LlmContentEncoder, LlmSseContentEncoder) forward from
#2571 so the collapse can happen where decode already lives, rather than
forking that package ahead of its own introduction there; #2571 will
need to rebase on top of this and drop its now-duplicate copies.

LlmSseContentDecoder, LlmSseContentEncoder, LlmJsonContentDecoder widen
from package-private to public (unchanged otherwise) since their new
LlmSseContentCodecSpi/LlmJsonContentCodecSpi providers construct them
from the sibling internal.codec package. Adds LlmJsonContentEncoder
(new): the application/json inverse of LlmJsonContentDecoder, copying
content bytes through unchanged with no framing on flush.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Mw32oxEw24fLt5Ypakj6pH
@claude
feat(binding-llm): LlmDialectFactorySpi / LlmDialect exported SPI
fcf279f 
Defines the pluggable-dialect contract for binding-llm, exported from the
start (unlike LlmContentDecoderSpi, which stays internal): LlmDialect
exposes name()/detect()/contentType() plus supplyDecoder(Kind)/
supplyEncoder(Kind) returning common-json JsonTransform stages, and
LlmDialectFactorySpi is the ServiceLoader-registered entry point.
HttpHeaders is a minimal read-only accessor for detect(path, headers),
since no HTTP header abstraction previously existed in this codebase and
pulling in jakarta.ws.rs would add a dependency never otherwise used here.
Kind is nested on LlmDialect, distinguishing request/response schemas.

No concrete dialect implementations yet (OpenAI/Anthropic land later) --
module-info.java exports the dialect package and declares uses without a
corresponding provides. Unit-tested via a stub LlmTestDialect/
LlmTestDialectFactorySpi registered under test-scope META-INF/services,
mirroring this module's existing LlmContentDecoderSpi/
LlmTestContentDecoderFactorySpi pattern.

Fixes #2480

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016NcAVwRPN1Pwjzpobr75w6
@claude
feat(binding-llm): resolve LlmDialect content-type per request, not p…
efda042 
…er instance

contentType() previously took no parameters, so a dialect could only
report one fixed content-type for its lifetime -- insufficient for an
API whose response framing (event-stream vs. a single JSON document)
depends on a flag in the request body, since neither contentType() nor
detect(String, HttpHeaders) offered any way to inspect it.

Adds HttpRequestBody, a minimal read-only scalar-member accessor
mirroring HttpHeaders, and changes contentType() to
contentType(Kind, HttpHeaders, HttpRequestBody): Kind lets request and
response resolve independently (a dialect's request body content-type
can be fixed while its response varies), and the headers/body context
lets that resolution depend on the actual request rather than being
fixed at dialect-instance-creation time. Both parameters are nullable
for callers without that context available.

LlmTestDialect now resolves text/test-event-stream for a streaming
response and application/test+json otherwise, exercising the new
per-Kind, per-request resolution the stub previously couldn't express.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016NcAVwRPN1Pwjzpobr75w6
@claude
refactor(binding-llm): resolve dialect detection/decoding via ModelEn…
22d3390 
…velope, not whole-body buffering

Replaces LlmDialect's HttpHeaders/HttpRequestBody/contentType() with the
engine's existing ModelEnvelope/ModelTransform (runtime/engine/.../model/),
reusing machinery this codebase already has instead of inventing
LLM-specific buffering to resolve text/event-stream vs. application/json,
or a model name, by peeking one field of an otherwise-unbuffered body.

detect(ModelEnvelope) folds :path/:method into the same envelope as
ordinary headers -- no separate path parameter. supplyDecoder/supplyEncoder
now take (Kind, ModelEnvelope) and return ModelTransform: a per-field
stage that can extract a field (e.g. a model name) into the envelope while
the body still flows through unchanged, mirroring
KafkaExtractTransform (runtime/binding-kafka/.../cache/) -- so a caller
reads that signal back off the envelope as decoding proceeds rather than
buffering the whole body first to inspect it. contentType() is removed
entirely: nothing in this shape needs it once the streaming/non-streaming
signal is just another envelope entry a caller reads after extraction.

common-json/JsonTransform is no longer used anywhere in this module now
that the dialect SPI itself doesn't need it, so the dependency comes back
out of module-info.java and pom.xml along with it.

LlmTestDialect/LlmTestDialectFactorySpi become LlmTestConditionalDialect/
LlmTestConditionalDialectFactorySpi, since what they now demonstrate is
exactly this: request detection and model-name extraction conditional on
envelope contents, not a fixed per-instance answer.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_016NcAVwRPN1Pwjzpobr75w6
@claude
feat(binding-llm.spec): canonical event vocabulary — choice index, to…
300c55f 
…ken detail

Extend LlmFlushEx's block-lifecycle skeleton (from #2476) with the two
places OpenAI's parallel completions and per-token detail need to survive
in the vocabulary, per the issue's scope:

- choiceIndex (default 0) on messageStart, blockStart, blockEnd, finish,
  and native/raw: the (choice, block) compound index's outer half.
  Anthropic is always choice 0, so every existing dialect mapping is
  unaffected; OpenAI's n > 1 becomes one messageStart per parallel
  completion, distinguished by choiceIndex. usage stays choiceIndex-free
  since every dialect reports it aggregated across choices, never per
  choice.
- logProbability (nullable) on LlmDataEx: the one per-delta detail the
  vocabulary carries directly, for dialects exposing per-token detail
  (OpenAI logprobs) without reopening the DATA/FLUSH split from #2476 or
  growing the vocabulary for the full log-probability structure — richer
  detail than one value per token stays behind LlmNativeFlushEx.

Documents the three lossiness cases as doc comments alongside the fields
they concern: choiceIndex and logProbability both drop out on any
cross-dialect route to Anthropic (structurally exactly one choice, no
per-token detail); message-start input token counts are resolved by
decoupling inputTokens into its own deferred usage event rather than
emitting a placeholder on messageStart and correcting it later, so a
source that discloses tokens late (OpenAI) just emits usage late instead
of needing a correction event.

Fixes #2481

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AYGZDrytLjmG2AQqUN1hsu
@claude
feat(binding-llm): event sequence mapper for anthropic and openai dia…
ff41aa1 
…lects

Translates between each dialect's native streaming event sequence and the
canonical vocabulary from #2557, in both directions:

- LlmAnthropicEventMapper: holds input_tokens from message_start until the
  paired usage event at message_delta; tracks the currently open block's
  type to know when content_block_stop needs a canonical blockEnd (tool
  calls only) and to route content_block_delta payloads (text_delta vs
  input_json_delta) on encode.
- LlmOpenAiEventMapper: translates OpenAI's tool-call-only index space into
  the canonical (Anthropic-shaped) block index via a per-stream map, and
  synthesizes blockEnd lazily -- deferred until the next tool call starts
  or the stream finishes, since OpenAI has no explicit block-close event.

Unit-tested against both worked-example tables from the issue (message
role/content/tool-call cardinality changes in each direction), plus the
held-usage/already-consumed and lazy-blockEnd edge cases.

Fixes #2482

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_019BHBkLjV2tcbNxMEgxkpSw
@claude
feat(binding-llm): dialect detection dispatcher and fixed-dialect config
53e53df 
LlmDialectResolver dispatches path/header detection across every LlmDialect
registered via LlmDialectFactorySpi, without hardcoding any dialect's
signals. A configured fixed dialect name bypasses detection entirely,
including when it matches no registered dialect. When detection matches
more than one dialect, or none, resolution is ambiguous and returns null
so the caller rejects the request rather than guessing.

LlmOptionsConfig adds the optional server-kind `dialect` option (schema,
config, and adapter) used to pin a fixed dialect.

Fixes #2483

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01AtkteS7C66qiVLGeEDJ2FP
@claude
feat(engine): pre-seed catalogs in the sys namespace patch contract
99f8666 
The sys: namespace's patch contract (Binding.system()/Exporter.system())
lets a component contribute a shared binding (e.g. http_client) but has
no pre-seeded slot for a shared catalog, so a patch adding one has
nothing to append into. Pre-seed catalogs: {} alongside the existing
bindings: {} in the base sys namespace skeleton, the same "pre-seed the
extension point" convention already used for the JSON-schema *-ext
scaffolds, so any component can share a catalog-backed resource the way
bindings are already shared.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
@claude
feat(binding-llm): generate a shared sys catalog from registered dial…
9b620c6 
…ect schemas

Adds LlmDialectFactorySpi.schema(Kind), letting a dialect contribute a URL
to its own JSON schema for the request or response direction without
knowing anything about catalogs or patches. LlmBinding.system() enumerates
every registered dialect, reads each contributed schema, and generates a
sys: namespace patch adding one shared inline catalog (llm_dialects) with
a <dialect>.request/<dialect>.response subject per schema a dialect
contributes -- built once, at engine startup, since the dialect set is
ServiceLoader-discovered off the classpath and therefore fixed for the
JVM's life, the same way sys: already shares a binding (e.g. http_client)
across every binding that references it.

LlmDataUrlStreamHandler decodes a base64 data: URL (RFC 2397) in memory,
scoped to a single URL via the URL.of(URI, URLStreamHandler) factory --
no temporary file, no globally-registered protocol handler -- used to
hand the generated patch to Binding.system()'s URL-returning contract
without writing it to disk.

LlmBeginEx gains contentType and model fields alongside dialect, for a
server-kind stream factory to stamp once it has resolved the dialect and
read the request's content-type.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
@jfallows
jfallows force-pushed the claude/stoic-sagan-9d81tf branch from 5f11258 to 9b620c6 Compare September 15, 2026 18:41
@claude
feat(binding-llm): decode network requests, forward to app0, annotate…
1f31d79 
… LlmBeginEx.dialect

Implements request/response framing decode-and-forward for the LLM server
binding (#2484): LlmServerFactory drives the request body
through the resolved dialect's ModelPipeline as bytes arrive off the wire,
forwards transformed content to app0 incrementally rather than buffering
the whole body, and threads the resolved dialect name onto LlmBeginEx so
app0 can see which wire dialect produced the request.

Flow control between the client, this binding, and app0 is enforced with
dedicated decodeSlot/encodeSlot buffers on each side of the exchange
(LlmServer for the network-facing leg, LlmStream for the app-facing leg),
each granting credit strictly from its own local slot occupancy rather than
copying a peer's sequence numbers across independent byte domains. LlmState
tracks per-direction open/closing/closed transitions and end-of-stream
deferral while a buffer is still draining.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
@claude
fix(binding-llm): give network-decode and app-relay slots separate Bu…
9898942 
…fferPool handles

DefaultBufferPool.buffer(slot) rewraps and returns a single shared mutable
field per pool instance, so holding a buffer reference across a nested call
that fetches a different slot from the same pool silently repoints it.
LlmServer.decodeNetwork() directly, synchronously calls LlmStream's
request-relay staging method as a plain nested Java method call, so a
single pool instance backing both the network-decode slot and the app0
request-relay slot would alias between them.

Give LlmServerFactory two BufferPool handles instead of one: decodePool
(network decode) and encodePool (app0 relay, both directions), obtained via
context.bufferPool() and .duplicate() -- matching McpServerFactory's
decodePool/encodePool precedent. The two relay directions sharing
encodePool (the reply-direction slot on LlmServer and the request-direction
slot on LlmStream) never appear in the same call stack: cross-binding
accept() is ring-buffer-mediated and dispatched on a later engine tick, not
a nested call, so one encodePool instance can't alias between them.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
@jfallows
jfallows force-pushed the claude/stoic-sagan-9d81tf branch from 3de04aa to 9898942 Compare September 16, 2026 00:02
@claude
fix(binding-llm): move catalog-inline to runtime scope, add catalog-i…
71332ef 
…nline.conf

LlmSystemNamespaceGenerator emits a real "type": "inline" catalog config
into the generated sys: namespace patch, serviced at actual runtime by
catalog-inline's CatalogFactorySpi via ServiceLoader -- not just exercised
by this module's own tests. test scope kept it off the runtime classpath
entirely; provided scope wouldn't fit either, since nothing in main source
compiles against catalog-inline's Java API (the reference is a plain
string), so there's nothing to satisfy at compile time.

Match binding-asyncapi/binding-openapi's precedent for this same
generated-"type: inline"-config pattern: catalog-inline at runtime scope,
plus the companion catalog-inline.conf (config-schema side) at default
scope alongside the existing model-json.conf dependency.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
@claude
fix(binding-llm): move model-json to runtime scope
6d1f276 
LlmBindingConfig.newModelConfig() constructs a real JsonModelConfig in main
source, resolved to a working ModelHandler by context.supplyModel() via a
ModelFactorySpi lookup at actual runtime -- not just exercised by this
module's own tests. Matches binding-asyncapi/binding-mcp-openapi/
binding-openapi, each of which also builds JsonModelConfig directly in main
source and declares model-json at runtime scope.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
@claude
fix(incubator): regenerate aggregate NOTICE
dee2717 
Stale relative to the catalog-inline/model-json scope fixes on binding-llm
(catalog-inline.conf and model-json.conf now roll up into this aggregate),
plus a pre-existing gap for binding-http.spec's license entry. Regenerated
via ./mvnw notice:generate -pl incubator -amd, never hand-edited, per
AGENTS.md.

Co-Authored-By: Claude Sonnet 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_015mHT1vrPATQGSdRcB8NNKM
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Reviewers

No reviews

Assignees

No one assigned

Labels

None yet

Projects

None yet

Milestone

No milestone

Development

Successfully merging this pull request may close these issues.

llm server: framing decode and forward, annotate LlmBeginEx.dialect

2 participants

@jfallows @claude