You signed in with another tab or window. Reload to refresh your session.You signed out in another tab or window. Reload to refresh your session.You switched accounts on another tab or window. Reload to refresh your session.Dismiss alert
Offline mock:// function tools emit OpenAI message.tool_calls and SSE delta.tool_calls / finish_reason=tool_calls.
Invoice lookup_balance binds INV- from the user text (default INV-9). Named tool_choice and multimodal text parts are honored; tool_choice=none stays content/stop.
APA 7th: OpenAI function calling; Schick et al. (2023) Toolformer; Yao et al. (2023) ReAct; Holtzman et al. (2020) nucleus sampling.
Buyer next action
Send stream=true on tool-calling requests and include the invoice id in the user message. On streamed route completions, send the nucleus and penalty values you want — they are no longer dropped.
Do not merge in parallel
Prefer this over #606 / #617 / #609 / #610 / #613. Keep KV/token #621 and message-name #623 on their own tips. Independent non-author APPROVE + Full unit/Semgrep required.
Local checks printed ok: SSE tools honesty, stream sampling honesty, related tools-passthrough honesty, paper/API/self-check/conventions.
…closed otherwise
Chat history: message-level audio and legacy function_call are null/empty omit
no-ops; non-empty fail closed with named errors (including tools passthrough).
Tip substrate from #577 assistant refusal/annotations honesty. Local full unit:
940 passed.
…ed otherwise
OpenAI fine-tune style message weight is not applied on this gateway.
Accept null/0/1 as honest no-ops; reject other types and values with
invalid_message_weight. Tip substrate from #578. Local full unit: 943 passed.
…ion role
Reject unsupported message keys with named unknown_message_fields (not silent
strip or tools-passthrough smuggle). Reject legacy function role with
invalid_message_role migration to tool. Tip substrate from #579. Local full
unit: 947 passed.
OpenAI partial-assistant prefix flag is not applied on this gateway.
null/false are honest no-ops; true and non-booleans fail closed with
invalid_message_prefix. Tip substrate from #580. Local full unit: 950 passed.
…therwise
Named invalid_max_tool_calls on /v1/chat/completions instead of opaque
unknown_fields. Aligns with Responses max_tool_calls honesty; gateway has
no multi-step tool loop.
…losed otherwise
Legacy /v1/completions treated max_tool_calls as unknown_fields. Accept the
key for named invalid_max_tool_calls (null/empty/whitespace omit-equivalent),
matching chat/Responses honesty so SDKs get a clear migration path.
SDK clients often send include_usage/include_obfuscation as JSON null.
Drop null flag values before validation so null (and null+false mixes)
match omit / all-false no-ops on chat, Completions, and Responses.
True flags remain fail-closed with invalid_stream_options.
…age honesty
Null include_usage/include_obfuscation stay omit-equivalent, but unknown
stream_options keys no longer become no-ops just because their value is
null. Weight, prefix, refusal, annotations, developer role, empty
user/system content, and participant name now use the same named errors
on the tools passthrough path as on orchestration.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Hoist stream, required model, stream_options, and temperature/top_p range
checks before proxy_completion so a tools or response_format body cannot
return a billed JSON completion when the SDK asked for SSE, or silently
pick a pool model when model is omitted.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Hoist stream, required model, stream_options, and temperature/top_p range
checks before proxy_completion so a tools or response_format body cannot
return a billed JSON completion when the SDK asked for SSE, or silently
pick a pool model when model is omitted.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Hoist attribution and routing validation before proxy_completion so a
tools or response_format body cannot bill a sync completion with an
unknown spend dimension or a batch/latency_tolerant hint. Tools
passthrough has no batch job plane.
Buyer next action: send known sync attribution; omit routing.channel=batch
and latency_tolerant=true on tool-calling requests.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Call the orchestration message, max_tokens, attribution, and routing
validators before proxy_completion so a tools or response_format body
cannot bill a completion with no prompt, crash on a non-object entry,
or silently drop unknown spend/routing keys.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Match _validate_messages: tools + user content 123 must 400 invalid_message
instead of billing a sync completion. Assistant/tool JSON null stays
omit-equivalent.
Buyer next action: send user/system content as a non-empty string or a
content-parts array.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Match the tools-path honesty contract to the invalid_message behavior
so SDK clients send a string or content-parts array.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Passthrough has no batch job plane. Reject routing.channel=batch and
latency_tolerant=true before proxy_completion so a tool-calling body
cannot bill a silent sync completion.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Hoist the remaining chat request knobs before proxy_completion so an
OpenAI SDK tool-calling body cannot bill a sync completion for seed,
stop, n>1, logprobs, logit_bias, out-of-range token/penalty values,
unsupported reasoning_effort, or a non-default service_tier.
Buyer next action: omit those fields on tool-calling requests; they are
not applied on this gateway.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Call _validate_messages before proxy_completion so tools + [] / omitted /
null / non-list / non-object messages raise invalid_message instead of
billing a completion with no prompt.
Buyer next action: always send a non-empty messages array of objects on
tool-calling requests.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Hoist the orchestration request-knob validators before proxy_completion
so a tools or response_format body cannot bill a completion that silently
drops seed, stop, n>1, logprobs, logit_bias, or out-of-range penalties.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
stream_chat defaulted to 0.2 even after the HTTP path wrote
default_temperature from the request. A streamed invoice summary at
temperature=0.8 no longer silently falls back to 0.2.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
… proxy
SDK optional defaults serialize omitted fields as JSON null. Accepting those
keys without popping them is not omit-equivalent: proxy_completion forwards
the body and several providers reject a null JSON Schema object.
Pop the keys in place, keep non-null wrong types on invalid_tools, and assert
omit-real via mock echo on chat and Responses.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
OpenAI SDK tool callers send stream=true. Returning 400 was honest but
left every streaming tool client failing. Pipe a single pool agent as
chat.completion.chunk frames so concatenated mock content matches the
non-stream JSON body, and pipe live provider text so tool_calls deltas
survive. Keep include_usage, empty messages, and unsupported knobs
fail-closed.
Buyer next action: send stream=true on tool-calling requests; omit
stream_options.include_usage.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Keep tools/response_format SSE proxy from this branch and take #601's
null tool.function omit-pop, request-knob HTTP cases, and streamed
route temperature. Conflict resolution keeps stream=true as SSE, not
400. CHANGELOG records the combined unreleased slice.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Unknown assistant tool_calls entry/function fields, non-boolean
include_orchestration_trace, and unknown mode used to bill a JSON or SSE
completion on the tools proxy. They now return the same named 400s as the
orchestration path before the first byte.
Buyer next action: send only id/type/function/optional index on assistant
tool_calls; send a boolean or omit include_orchestration_trace; send
auto/route/conduct or omit mode.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
mock:// lookup_balance now returns assistant tool_calls and streams
delta.tool_calls with finish_reason=tool_calls so offline SDK clients
match the live _stream_raw path. Invoice identifiers bind from the
user text (default INV-9). Cite Toolformer, ReAct, and OpenAI function
calling (APA 7th) in docs/papers and the honesty test.
Buyer next action: send stream=true on tool-calling requests and include
the invoice id in the user message.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
stream_chat recorded temperature but omitted the request-scoped top_p,
presence_penalty, and frequency_penalty that chat() already forwards.
A streamed invoice summary at top_p=0.1 no longer drops those knobs.
Buyer next action: send the nucleus and penalty values you want on
stream=true route requests; streaming matches the JSON path.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
Streamed top_p / penalties now match chat(). Record the ICLR 2020
nucleus-sampling paper (APA 7th) so the buyer can see why those knobs
must not silently drop on stream=true.
Buyer next action: send the nucleus and penalty values you want on
streamed route requests.
Co-authored-by: Seongho Bae <seonghobae@users.noreply.github.com>
seonghobae
changed the title
fix(api): mock SSE tool_calls for streamed function tools
fix(api): mock SSE tool_calls and streamed route sampling
Aug 16, 2026
Scoped to the three unique commits (8d7d534, a50032e, baf2e84), not the stacked 154-file PR vs main. CodeRabbit CLI 0.7.3 is installed here but not authenticated, so this is a HEAD-file + unique-diff review. Local tests/test_passthrough_sse_tools_http_honesty.py and tests/test_stream_chat_request_sampling_http_honesty.py both printed ok.
What holds
Offline mock:// function tools emit JSON message.tool_calls / finish_reason=tool_calls with content: null. _mock_raw_sse emits delta.tool_calls (name frame, then 24-char argument chunks) and finishes as tool_calls.
Invoice lookup_balance binds INV- from user text, including multimodal text parts; missing id defaults to INV-9. Named tool_choice selects that function. tool_choice=none stays content/stop.
Streamed route top_p / presence_penalty / frequency_penalty read the same request-scoped ModelClient defaults chat() already uses. HTTP tests prove the server writes those defaults before stream_route.
No naming-rule, KV-not-env, or security defect in this slice. Holtzman et al. (2020) is cited APA 7th in CHANGELOG.md and docs/papers/README.md. Known residual /v1/responses streaming stays on #628 — do not fold it here.
Findings: 0 Critical, 0 Warning, 3 Info
Info-only residuals are inline. They do not block the unique slice.
Buyer next action
Send stream=true on tool-calling requests and put the invoice id in the user text. On streamed route completions, send the nucleus and penalty values you want — they are no longer dropped. Independent non-author APPROVE plus Full unit/Semgrep are still required before merge. Do not merge #606 / #617 / #609 / #610 / #613 in parallel with this tip. Keep KV/token #621 and message-name #623 on their own tips.
Checks were still queued at review time; queued GitHub Checks are not treated as a defect in this slice. This automation cannot approve or merge.
Unique slice vs #606 (0717921) is SOUND for official-SDK payloads. Mock lookup_balance emits JSON message.tool_calls and SSE delta.tool_calls / finish_reason=tool_calls. Exact tool_choice=none stays content/stop. Streamed route top_p / penalties match chat(). Local honesty + paper/API/self-check/conventions printed ok.
Medium (blocking for padded-none honesty):_selected_function_name uses exact == "none" while _validate_chat_tool_choice already accepts " none " / "\tnone " and discards the stripped token. Repro on this tip: tools + tool_choice=" none " returns finish_reason=tool_calls after HTTP 200. Official SDKs send exact none and are fine.
Fix is on the successor tip (1550324): strip in the selector, write the token back on the body, and cover exact + padded none on JSON and SSE. Prefer that PR over merging this head alone.
Buyer next action: send exact tool_choice=none when you want no tool call; include the invoice id in the user text; omit stream_options.include_usage. Independent non-author APPROVE + Full unit/Semgrep still required. This automation cannot merge or self-approve.
The reason will be displayed to describe this comment to others. Learn more.
Medium: exact == "none" misses the padded values _validate_chat_tool_choice already accepts as none (" none ", "\tnone "). The validator strips for the type check but does not write the stripped string back, so passthrough still delivers the padded value here.
Repro on baf2e84: proxy_completion with lookup_balance + tool_choice=" none " (or "\tnone ") returns finish_reason=tool_calls. Exact "none" stays content/stop as claimed.
Fix on the successor tip: if isinstance(tool_choice, str) and tool_choice.strip() == "none": return None, and persist the stripped token on the request body so live providers see none too.
Deferring to tip #691 for product-gate focus (Full unit + Semgrep). Reopen if this delta is not on tip substrate.
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
Add this suggestion to a batch that can be applied as a single commit.This suggestion is invalid because no changes were made to the code.Suggestions cannot be applied while the pull request is closed.Suggestions cannot be applied while viewing a subset of changes.Only one suggestion per line can be applied in a batch.Add this suggestion to a batch that can be applied as a single commit.Applying suggestions on deleted lines is not supported.You must change the existing code in this line in order to create a valid suggestion.Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be applied on multi-line comments.Suggestions cannot be applied while the pull request is queued to merge.Suggestion cannot be applied right now. Please check back later.
seonghobae
changed the title
![cursor[bot]](https://avatars.githubusercontent.com/in/1210556?s=60&v=4){kind=small}
Landing vehicle for the remaining #606 buyer gap, plus the #617 sampling slice.
Unique vs #606 (
0717921)mock://function tools emit OpenAImessage.tool_callsand SSEdelta.tool_calls/finish_reason=tool_calls.lookup_balancebindsINV-from the user text (defaultINV-9). Namedtool_choiceand multimodal text parts are honored;tool_choice=nonestays content/stop.top_p/presence_penalty/frequency_penaltymatchchat()(a1823cccherry-pick).Buyer next action
Send
stream=trueon tool-calling requests and include the invoice id in the user message. On streamed route completions, send the nucleus and penalty values you want — they are no longer dropped.Do not merge in parallel
Prefer this over #606 / #617 / #609 / #610 / #613. Keep KV/token #621 and message-name #623 on their own tips. Independent non-author APPROVE + Full unit/Semgrep required.
Local checks printed
ok: SSE tools honesty, stream sampling honesty, related tools-passthrough honesty, paper/API/self-check/conventions.