Expose max in the Codex thinking dropdown for gpt-5.6-sol only (maps to
xhigh on wire; live probe rejected ultra). Include custom/kilo models
when computing provider thinking options so manually added gpt-5.6-sol
contributes its max level.
Co-authored-by: Cursor <cursoragent@cursor.com>
Translator only unwrapped tool.function when both tool.type==="function"
and tool.function were truthy. Loose/legacy OpenAI clients emit the bare
{ function: { name, parameters } } shape (no parent type), which fell
through and forwarded name: undefined upstream, rejected by strict
Anthropic-compatible gateways (MiniMax M3) as (2013) invalid tool type.
Unwrap tool.function whenever present. Built-in tools stay pass-through.
Adds regression coverage for the 4 tool shapes. See #2435.
Co-authored-by: Cursor <cursoragent@cursor.com>
Claude Code sends reasoning_effort "max" (its top level); OpenAI enum caps
at "xhigh" and rejects "max" with HTTP 400 "max effort not support".
applyFormat case "openai" now clamps "max"->"xhigh" before assigning
body.reasoning_effort; other levels pass through unchanged.
Add regression test covering client output_config.effort, direct
reasoning_effort, passthrough of xhigh/high, and budget_tokens capping.
Co-authored-by: Cursor <cursoragent@cursor.com>
isPathLike rejected any line with a colon, so Windows absolute paths
(C:\Users\me\a.js) were never recognized and find dumps went uncompacted.
find.js also split only on "/", mis-grouping backslash paths.
- autodetect: treat drive-letter prefix (X:\ or X:/) as path-like before
the general colon rejection.
- find.js: split on the last "/" or "\" separator and normalize emitted
directory labels to forward slashes.
Co-authored-by: Cursor <cursoragent@cursor.com>
- map service_tier=fast to upstream priority; drop unsupported tiers
- normalize reasoning effort max to xhigh (codex-only)
- convert 200-SSE model-capacity errors into 503 so account fallback rotates
- keep normal SSE output intact after peeking
Co-authored-by: Cursor <cursoragent@cursor.com>
VolcEngine Ark caps the Kimi family at max_tokens <= 32768, but the
model's advertised ceiling is far higher (Kimi-K2.7-Code resolves to
maxOutput 262144), so clampToModelMaxOutput alone leaves it uncapped and
the request 400s. Add a Kimi-scoped rule with an explicit maxOutputCap of
32768, combined with the model ceiling via min(). Covers max_tokens,
max_completion_tokens, max_output_tokens.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Match captured official Antigravity IDE traffic: cloudcode-pa host,
antigravity/ide/2.1.1 User-Agent, IDE-shaped agent requestId, and drop
router-only stream/usage headers plus the legacy double system prompt.
Co-authored-by: Cursor <cursoragent@cursor.com>
Ark rejects max_tokens above 128000 for GLM-5.2. Add a config-driven STRIP_RULES entry that clamps max_tokens, max_completion_tokens and max_output_tokens down to the model maxOutput before the upstream call.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add four shared Caveman prompt fragments (no invented abbreviations,
preserve user language, no self-reference, no decoration) across all six
levels, and remove ULTRA contradictions around abbreviations/arrow
shorthand. Adds regression tests for the prompt rules.
Co-authored-by: Cursor <cursoragent@cursor.com>
Map role="developer" messages to top-level instructions alongside
role="system" in openaiToOpenAIResponsesRequest. Previously developer
messages matched no branch and were silently dropped from the Responses
request, losing GPT-5/Codex system-level prompts.
Co-authored-by: Cursor <cursoragent@cursor.com>
On the translated OpenAI->Claude path, adjustMaxTokens capped max_tokens
before applyThinking set thinking.budget_tokens, so max-effort budget
(128000) could exceed a 64k-clamped max_tokens -> Anthropic 400.
prepareClaudeRequest now reconciles after the budget is known: prefer
raising max_tokens, only shrink budget when it meets/exceeds the ceiling.
Also lift the global 64000 cap: the ceiling is now the model's real
maxOutput, so high-output models (fable/mythos, opus-4.8/sonnet-4.6) get
their full budget. adjustMaxTokens gains an optional ceiling arg (default
unchanged, callers untouched); openai-to-claude passes the model maxOutput.
Native Claude Code passthrough is unaffected.
Co-Authored-By: Claude <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
- Pass system prompt via native systemInstruction field (+ <instructions> fallback)
so Claude models stop treating it as info-only <system-reminder>
- Add Opus 4.5/4.7/4.8 (base/thinking/agentic/thinking+agentic) to Kiro registry
- normalizeModelId(): dash->dot version separator, scoped to Kiro provider only
- Replace <system-reminder> with <instructions> in claude-to-openai/openai-to-kiro
Co-authored-by: Cursor <cursoragent@cursor.com>
Normalize every provider to one cache-inclusive convention via
canonicalizeUsage() before persist, and price cached + cache_creation as
subsets of prompt_tokens in calculateCostFromTokens() to stop
double-counting. usageRepo now delegates cost math to a single source.
Surface Cached tokens/cost across dashboard (overview, tokens, cost,
details). Merge Claude message_start cache with message_delta output so
cache counts survive. Compatible LLM nodes now allow multiple API-key
connections (key pool).
Co-authored-by: Cursor <cursoragent@cursor.com>
handleStreamingResponse and buildOnStreamComplete each generated their
own streamDetailId for what should be one logical record — the
placeholder row (0 tokens) and the final row (real usage) never shared
an id, so the DB's ON CONFLICT(id) upsert never merged them, leaving a
permanent 0-token stub for every streaming request.
Share the id from buildOnStreamComplete with handleStreamingResponse
so both writes hit the same row.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add read-only GET to inspect per-credit reset inventory (status, granted,
expiry, remaining) with a Quota Tracker modal. DRY the route via shared
connection/refresh helpers; keep existing consume POST unchanged.
Co-authored-by: Cursor <cursoragent@cursor.com>
Combo mixes models, so non-Claude thinking signatures leak into
conversation history. Native passthrough forwarded them verbatim and
Anthropic rejected the request. Validate signatures and drop invalid
thinking blocks, re-inserting a placeholder when tool_use requires one.
Co-authored-by: Cursor <cursoragent@cursor.com>
- Updated capabilities for NVIDIA models to enforce OpenAI-compatible reasoning formats.
- Added new models: MiniMax M3, GLM 5.2, DeepSeek V4 Pro, DeepSeek V4 Flash, Kimi K2.6, and Nemotron 3 Ultra to the NVIDIA registry.
This enhances the provider's functionality and aligns with OpenAI standards.
IAM Identity Center (authMethod=idc) tokens failed every request with 403
"bearer token invalid". Treat idc like api_key/external_idp:
- executors/kiro.js: route idc to *.amazonaws.com CodeWhisperer surface,
region-aware from credentials.region instead of hardcoded us-east-1.
- openai-to-kiro.js / claude-to-kiro.js: send resolved profileArn or empty
for idc/external_idp, never the shared builder-id placeholder ARN.
Co-authored-by: Cursor <cursoragent@cursor.com>
Filter empty text blocks from thoughtSignature-only parts, preserve
tool_calls when functionResponse and functionCall coexist in the same
content, and skip empty regular text parts before they reach Claude.
Co-authored-by: Cursor <cursoragent@cursor.com>
Clients echo full message history each turn including reasoning_content,
which the Kimchi OpenAI gateway counts as input tokens. Multi-turn convos
balloon to 100k+ tokens and the model returns empty content.
KimchiExecutor.transformRequest now strips reasoning_content from assistant
messages when it exceeds an 8-char threshold, preserving the 1-char
placeholder injectReasoningContent sets and keeping content intact.
Co-authored-by: Cursor <cursoragent@cursor.com>
CodeBuddy CN includes "tool_calls": [] in every SSE streaming delta.
@ai-sdk/openai-compatible checks delta.tool_calls != null — an empty
array passes ([] != null is true in JS), triggering premature
reasoning-end on every reasoning chunk (0/1ms durations in OpenCode).
Strip empty tool_calls arrays in passthrough before hasValuableContent.
Zero side-effect: real tool_calls always have at least one element.
Closes#2176
Co-authored-by: Cursor <cursoragent@cursor.com>
Gemini response translation wrote OpenAI-shaped bookkeeping into the
shared state.toolCalls map, which the downstream openai-to-claude
translator uses for Claude block metadata. That pre-population skipped
blockIndex creation, so Anthropic input_json_delta events lost index.
Track Gemini function calls via state.geminiToolCallCount instead,
leaving state.toolCalls clean for the Claude translator.
Closes#2218
Co-authored-by: Cursor <cursoragent@cursor.com>
Treat response.done as a terminal OpenAI Responses stream event so
passthrough streams ending with response.done are not flagged incomplete
and no synthetic response.failed is emitted. Restore the data: [DONE]
sentinel for same-format Responses passthrough streams.
Co-authored-by: Cursor <cursoragent@cursor.com>
Guard openai-responses compression: skip Headroom when body.input
contains non-message items (function_call, function_call_output,
reasoning) to preserve the Responses contract instead of collapsing
them into chat messages.
Co-authored-by: Cursor <cursoragent@cursor.com>
Claude Code chèn role:system cuối messages[], trước đây bị map thành assistant
khiến hội thoại không kết thúc bằng user → provider OpenAI-compat (LiteLLM)
dịch ngược Anthropic trả 400 "assistant message prefill". Map system -> user
và wrap <system-reminder> để giữ ngữ nghĩa instruction.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add qwen omni (audio/video input), qwen3.5/3.6/3.7 (native vision/video),
and mark qwen coder & max as text-only reasoning models.
Co-authored-by: Cursor <cursoragent@cursor.com>
Merge adjacent same-role blocks and strip empty parts before sending to
Gemini, avoiding 400 INVALID_ARGUMENT on consecutive same-role messages.
Co-authored-by: Cursor <cursoragent@cursor.com>
Opt-in quirk preserveCacheControl keeps cache_control on content blocks
for alicode/alicode-intl, enabling DashScope prompt caching. signature
is always stripped; all other providers unchanged.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add Kimchi as a browser-token OAuth provider routed through its
OpenAI-compatible gateway. Discover live models for /v1/models and
provider models, normalize Claude-compatible requests, and wire up
provider connection tests.
Co-authored-by: Cursor <cursoragent@cursor.com>
Backfill DEFAULT_THINKING_AG_SIGNATURE onto functionCall parts missing it
(client history replay) and on Claude tool_use blocks, fixing 400
INVALID_ARGUMENT from Gemini-family APIs. Suppress the OpenAI-style
data: [DONE] sentinel for antigravity/gemini/vertex to avoid parser crashes.
Fixes#2193.
Co-authored-by: Cursor <cursoragent@cursor.com>
Gemini rejects the non-standard 'deprecated' keyword in nested tool
schemas with INVALID_ARGUMENT (400). Add it to UNSUPPORTED_SCHEMA_CONSTRAINTS
alongside 'optional' so it gets stripped during translation.
Co-authored-by: Cursor <cursoragent@cursor.com>
CodeBuddy CN bonus packs ("Bonus Pack N") are one-shot credits whose
CycleEndTime equals DeductionEndTime — they expire for good and never
replenish. The dashboard rendered their resetAt as "Reset in Xd",
implying a monthly refill.
Tag bonus packs recurring:false (refill packs recurring:true) in the
usage handler, forward the flag through parseQuotaData, and word the
quota table / progress bar as "Expires in" / "Expires at" for
one-shot packs.
CodeWhisperer leaks literal <thinking> blocks into assistantResponseEvent,
duplicating reasoning already routed via reasoningContentEvent. Track
inThinking state to strip these tags during SSE transform, handling split
chunks across tag boundaries.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add modelsFetcher + passthroughModels so the dynamic Kilo Gateway
catalog surfaces in the combo model picker, matching openrouter.js.
Co-authored-by: Cursor <cursoragent@cursor.com>
Gemini/Antigravity generateContent rejects the JSON Schema annotation
keywords deprecated, readOnly, writeOnly with a 400 INVALID_ARGUMENT.
MCP tool schemas (e.g. Claude Code) commonly set deprecated:true, making
every request with such a tool fail. Add them to
UNSUPPORTED_SCHEMA_CONSTRAINTS so cleanJSONSchemaForAntigravity removes
them recursively before the request is sent.
Co-authored-by: Cursor <cursoragent@cursor.com>
Cloudflare AI rule only sets flattenContent. Treat missing match as
provider-wide and missing drop as empty list to avoid crash. Fixes#1960.
Co-authored-by: Cursor <cursoragent@cursor.com>
Retry short-lived 5xx/capacity errors (500/502/503/504 + message
patterns) with bounded backoff capped at 15s; honor Retry-After/reset
hints and skip when wait is too long. Keep 400 non-retryable. Enable the
retry hook for 500 alongside existing 429/503.
Deduplicate sanitized Antigravity tool names before emitting the single
functionDeclarations group to avoid upstream "Tool names must be unique"
rejections.
Add Headroom size diagnostics and phantom-savings warning when reported
token delta does not shrink the outbound payload.
Co-authored-by: Cursor <cursoragent@cursor.com>
Carry Claude reasoning_effort/reasoning into OpenAI Chat, map into
OpenAI Responses reasoning.effort, and keep request-level effort
(incl. xhigh) across tool-result turns instead of collapsing to high.
Co-authored-by: Cursor <cursoragent@cursor.com>
Registry exposes the dashed id claude-opus-4-7; matchPattern treats "."
as a literal, so it missed the dotted pattern and fell through to the
generic claude opus entry (200k / claude-budget). Add an exact entry so
it resolves to 1M context + adaptive thinking, plus a unit test covering
the dashed Opus ids.
Co-authored-by: Cursor <cursoragent@cursor.com>
OpenAI-compatible apikey provider (chat/embedding/image) with dynamic
model discovery via modelsFetcher + passthroughModels. No executor needed.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add 1M context capability overrides for claude-opus-4.8 and -thinking
variants, and use the resolved capability contextWindow (fallback 200k)
instead of the hardcoded 200k estimate in the Kiro executor.
Co-authored-by: Cursor <cursoragent@cursor.com>
Codex (openai-responses) body.input holds Responses items, not OpenAI
messages. Translate input -> OpenAI -> compress -> back to input so the
Responses contract is preserved. Fixes#1998
Co-authored-by: Cursor <cursoragent@cursor.com>
Forcing reasoning_effort:"medium" + reasoning_summary:"auto" on plain
requests tripped CodeBuddy's content filter and returned an error (#2071).
Make reasoning params opt-in: only set reasoning_summary when the client
sent an explicit reasoning_effort; none/off still drops it.
Fixes#2071
Co-authored-by: Cursor <cursoragent@cursor.com>
Pass Gemini AUDIO/TTS generateContent requests through to Google's native
v1beta endpoint instead of converting to chat, with per-credential fallback
(504 timeout, 502 fetch failure). Accept client keys from Bearer,
x-goog-api-key, or ?key= while forwarding only the configured Gemini
credential upstream. Expose native v1beta model names and rewrites, and add
Gemini 3.1 Flash TTS to the catalogs.
Co-authored-by: Cursor <cursoragent@cursor.com>
Fetch the live model list from the Copilot /models endpoint and surface
it through /v1/models, replacing the hardcoded github entry so newly
shipped models appear without a code change. Catalog is cached per
credential and the Copilot token is refreshed on 401/403 before retry.
Also raise the connectivity-test budget to max_tokens:16, since Claude on
Copilot emits no choices at max_tokens:1 and produced a false negative.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add "text" to Codex Responses API allowlist so text.format.json_schema
reaches upstream for structured outputs instead of being stripped.
Co-authored-by: Cursor <cursoragent@cursor.com>
Keep provider-required streaming when client prefers JSON. The
Accept: application/json branch no longer flips stream back to false
for forceStream providers, fixing 400 errors on stream-only providers
(e.g. Command Code) for Hermes / Claude Code / other JSON clients.
Fixes#2031
Co-authored-by: Cursor <cursoragent@cursor.com>
Resolve Gemini TTS models from shared TTS catalog and provider registry
with a safe fallback, fixing requests resolving to models/undefined when
ttsConfig.models is empty. Add gemini-3.1-flash-tts-preview to catalogs.
Co-authored-by: Cursor <cursoragent@cursor.com>
logUsage now only logs to console; DB write removed. Streaming usage is
recorded once via saveUsageStats (onStreamComplete), eliminating duplicate
usageHistory rows that inflated dashboard totals.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add image generation for Antigravity provider via gemini-3.1-flash-image
and gemini-3-pro-image, exposed through Text to Image UI and
/v1/images/generations.
- registry: serviceKinds ['llm','image'] + image model entries
- executor: image model detection + image_gen request envelope
- chatCore: force stream=false for image models (generateContent)
- nonStreamingHandler: parse inlineData -> markdown image
- imageGenerationCore: useExecutor fast-path for executor delegation
- imageProviders/antigravity: image adapter with image input support
- usage/google: image models in quota whitelist
Co-authored-by: Cursor <cursoragent@cursor.com>
The pattern matcher marked *minimax-m3* as vision: false, causing
9Router to strip image attachments before forwarding upstream. This
broke Claude Code / Cursor / Cline vision flows when routing through
MiniMax-M3.
Scoped vision: true to *minimax-m3* only. M2.7 and the catch-all
*minimax* pattern remain vision: false: those models are text-only
(per MiniMax docs / NVIDIA NIM model card), so forcing vision there
would send images to a model that errors instead of degrading.
Co-authored-by: Cursor <cursoragent@cursor.com>
Fixes#1933 — upstream returns 403 "Illegal access" on the chat endpoint
when requests lack a browser-like User-Agent. Mirror OmniRoute mimocode
executor: rotate across 3 Chrome UA strings on both bootstrap and chat.
Co-authored-by: Cursor <cursoragent@cursor.com>
Workers AI rejects OpenAI content-part array shape; flatten text parts
to a plain string per message before sending.
Co-authored-by: Cursor <cursoragent@cursor.com>
Strip `type` field and fold `function.{name,description,parameters}` into
top-level {name, description, input_schema} before forwarding to Claude-format
endpoints. MiniMax (and other Anthropic-compatible providers) reject tools
carrying a `type` field with error code 2013 ("invalid tool type").
Refs #1939
Co-authored-by: Cursor <cursoragent@cursor.com>
Two additions on top of the merged CodeBuddy CN OAuth provider:
1. API key auth — connect with a direct API key (sent as
Authorization: Bearer), not only via OAuth device-code flow.
- registry: add authModes ["oauth","apikey"] + hasOAuth; combined Bearer
auth already forwards the key, token-refresh skips key connections.
- providers POST: accept dual-auth providers (authModes includes
"apikey") that live under category "oauth" — previously rejected as
"Invalid provider". Also fixes the same latent gap for xai.
2. Quota tracker — surface CodeBuddy CN credit balance on the usage
dashboard for both OAuth and API-key connections.
- registry: add transport.usage.url (Tencent billing endpoint) +
features.usage/usageApikey so the connection is quota-eligible.
- new CN-scoped handler services/usage/codebuddy-cn.js: POST the billing
meter endpoint, unwrap data.Response.Data.Accounts[]. The payload mixes
two credit types that must not be merged:
* refill/base ("基础体验包") — recurring allowance; cycle resets well
before the resource expires (CycleEndTime << DeductionEndTime).
Reads the *Cycle* balance, resetAt = next refresh. Cadence-labelled.
* bonus ("活动赠送包") — one-shot credits that expire at CycleEndTime.
Reads the plain Capacity balance. Labelled "Bonus Pack N".
One quota row per package, soonest-expiring first.
- register handler under "codebuddy-cn" in USAGE_HANDLERS.
Frontend needs no change — USAGE_SUPPORTED_PROVIDERS/USAGE_APIKEY_PROVIDERS
and the generic parseQuotaData branch already cover this shape.
Co-authored-by: Cursor <cursoragent@cursor.com>
codebuddy-cn shipped without a short alias, making its model prefix the
longest of any provider (codebuddy-cn/glm-5.2). Add alias + uiAlias
"cbcn" so the prefix becomes cbcn/glm-5.2, consistent with qoder (qd),
kiro (kr), codex (cx). "cbcn" = CodeBuddy CN; reserves "cbai" for a
future codebuddy-ai (intl) provider.
Backward-compatible: full id codebuddy-cn/... still resolves (resolver
matches both id and alias).
Also regenerate tests/__baseline__/alias-baseline.json: idToAlias entry
codebuddy-cn now maps to cbcn, and cbcn joins the sorted modelKeys list.
Co-authored-by: Cursor <cursoragent@cursor.com>
Remove "pattern" from UNSUPPORTED_SCHEMA_CONSTRAINTS so glob/grep tool
schemas keep the required "pattern" property for the antigravity provider.
Fixesdecolua/9router#1368
Co-authored-by: Cursor <cursoragent@cursor.com>
flattenToolHistory only recognized OpenAI-style tool calls. Anthropic-compatible clients (Claude Code, /v1/messages) send tool invocations and results as tool_use/tool_result blocks inside the message content array. Since panel calls strip the tools definitions (#1859), the panel expert models received structured tool history without schemas, causing them to fail or misbehave (leading to empty responses and 503 errors).
Extend flattenToolHistory to recognize and flatten tool_use and tool_result blocks in the content array into prose text, keeping panel expert execution robust and independent of the client API format.
PR decolua/9router#1910
Co-authored-by: Cursor <cursoragent@cursor.com>
Update OpenCode Go model catalog to match the official Go docs.
Route Qwen 3.7 and MiniMax models through /v1/messages with
Anthropic-compatible headers instead of OpenAI-compatible chat.
Keep GLM, Kimi, DeepSeek, and MiMo on /chat/completions; use
kimi-k2.7-code because the live Go API rejects kimi-k2.7 for chat
while the docs config example uses the code model ID.
PR #1931 by nguyenha935
Co-authored-by: Cursor <cursoragent@cursor.com>
Build on the optional Headroom Token Saver from Carmelo Campos
(PR: feat: add optional Headroom token saver). Add managed start/stop
of the local headroom proxy from the dashboard, install detection,
status probing, and a simplified Token Saver UI.
- detect headroom CLI + python>=3.10, probe proxy /health
- spawn/stop proxy as a detached, pid-tracked process
- /api/headroom/{status,start,stop} routes, gated local-only in dashboardGuard
- one-click Start/Stop Headroom modal, no manual config needed
- claude<->openai shape conversion for /v1/compress via 9router translators
Thanks to Carmelo Campos (@carmelogunsroses) for the original Headroom integration.
Co-authored-by: Cursor <cursoragent@cursor.com>
## Fixes
- **Kiro**: honor thinking effort budgets
- **AG/Kiro/Xiaomi**: provider fixes
- **Combo/Fusion**: flatten tool history in panel calls to prevent 503
- **LLM selector**: show custom vision models in selector and model list
- **Image**: prevent compatible nodes from shadowing provider aliases
Resolve the Kiro thinking budget from client intent (OpenAI reasoning_effort,
OpenAI Responses reasoning.effort, Claude output_config.effort, and Claude
thinking.budget_tokens) by reusing the shared thinkingUnified extractThinking
parser, then inject the resolved budget into the Kiro thinking system prefix.
Explicit none/off/disabled stops the prefix injection; synthetic -thinking
aliases keep the default budget.
Co-authored-by: Cursor <cursoragent@cursor.com>
Panel models in the fusion strategy must answer in prose. When the request
carried tools or prior tool_calls/tool messages, agentic panel models kept
emitting tool_calls instead of prose, so extractPanelText() returned empty
and the engine fell into the 503 "All fusion panel models failed" branch.
Panel fan-out now strips tools/tool_choice and flattens tool turns into
assistant prose (instead of dropping them), so panels keep the context but
cannot loop on tools. The judge still receives the unmodified history.
Co-authored-by: warelik <warelik@WARELIK-MB.local>
Co-authored-by: Cursor <cursoragent@cursor.com>
Expose user-added imageToText custom models as vision-capable chat
models in the default LLM selector and /v1/models, map custom service
kinds to runtime capabilities, and keep typed filtering for
/v1/models/{kind}.
Co-authored-by: Cursor <cursoragent@cursor.com>
Add config-driven stripUnsupportedParams helper and use it in the default and github executors. Removes the deprecated temperature param for claude-opus-4 models (Anthropic 400) and consolidates github's scattered capability checks into one rule table.
Fixes#1748
Co-Authored-By: fjia <fjia@suntekcorps.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Adds Fusion as a third combo strategy alongside fallback/round-robin. A
fusion combo fans the prompt out to all member models in parallel, then a
configurable judge model synthesizes one final answer from the panel.
- handleFusionChat in open-sse/services/combo.js: quorum-grace collection
caps the straggler penalty, anonymized sources prevent judge brand-bias,
degrades to a direct answer on single survivor and 503 on total failure.
- chat.js dispatches strategy==="fusion" at both combo entry points.
- Combos dashboard: per-combo strategy Select replaces the round-robin
toggle, fusion reveals a judge picker, plus a strategy/capacity explainer.
- tests/unit/combo-fusion.test.js covers fan-out, judge routing/default,
quorum-grace straggler drop, single-survivor and total-failure degradation.
Co-authored-by: Cursor <cursoragent@cursor.com>
Adds long-lived API-key (ksk_) authentication for Kiro/AWS CodeWhisperer
and a direct claude:kiro / kiro:claude translation route that avoids the
lossy OpenAI two-hop pivot.
- translator: claude-to-kiro request + kiro-to-claude response translators,
registered on the exact source:target pair (direct route ahead of the
OpenAI pivot in index.js). claude-to-kiro uses shared schema constants
(ROLE/CLAUDE_BLOCK/DEFAULT_IMAGE_MIME) per app convention.
- auth: POST /api/oauth/kiro/api-key imports + validates a key via
ListAvailableProfiles, persists authMethod="api_key" (no refresh token).
- executor: send tokentype: API_KEY header and try *.amazonaws.com hosts
first for api-key creds; OAuth keeps kiro.dev first.
- fix: never inject the default placeholder profileArn for api-key auth
(CodeWhisperer 403s an ARN not owned by the key's account).
- ui: API Key method in the Kiro connect modal; surface api-key accounts
on the Quota Tracker and provider count.
- stream: env-overridable TTFT vs stall timeouts + Kiro keepalive frame.
- tests: claude-kiro-direct + kiro-profile-arn (11 tests).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Co-authored-by: Cursor <cursoragent@cursor.com>
Some providers (e.g. xiaomi-tokenplan -claude models) return OpenAI-format
responses even when request was translated to Claude. Early-return now detects
choices[]. Also strip reasoning_content only when content is non-empty so
thinking models keep their only output.
Closes#1836
Co-authored-by: Cursor <cursoragent@cursor.com>
