cliproxyapi
sloproxy
Star
Fork
Watch

Commits on main

nekomimi.pet
fix(claude): stop cloaking instructions as conversation turns
committed
nekomimi.pet
fix(claude): keep rebuilt system instructions top-level
committed
nekomimi.pet
merge upstream main with local compatibility fixes
committed
Luis Pater
chore(codex): update codex user-agent to 0.153.3
committed
sususu
feat(usage): propagate session and parent session hierarchy to usage reporting queue
committed
Luis Pater
fix(codex): scope usage limit errors to credentials and parse flexible quota resets
committed
Luis Pater
fix(claude): avoid mid-conversation system splicing for advisor calls or results
committed
Luis Pater
fix(watcher): preserve concurrent file updates during auth snapshot rescans
committed
Luis Pater
fix(aistudio): normalize thinking level to uppercase
committed
Luis Pater
chore(models): remove gpt-5.4 and gpt-5.4-mini models
committed
Luis Pater
fix(codex): preserve empty supported_reasoning_levels array
committed
Luis Pater
feat(models): add gpt-6-astra model and update codex client configurations
committed
rome-xi
perf(antigravity): batch reasoning replay mutations
committed
sususu
fix(gemini): ensure functionResponse normalizes to user role in Gemini request normalizer
committed
sususu
fix(antigravity): use ContentHasGeminiFunctionResponse instead of gjson projection
committed
sususu
fix(translator): flush pending developer notice before intervening user turn
committed
sususu
fix(translator): preserve Gemini prompt cache by demoting mid-session developer messages (#5490)
committed
sususu
feat(claude): harden probe and helper request classification, diagnostics isolation, and late cloaking
committed
sususu
feat(claude): add Fable 5.1 reporting outcomes block and post-payload reconciliation
committed
sususu
feat(claude): add 2.1.258 dynamic beta headers, model fallbacks, and paired cache TTL
committed
sususu
feat(claude): implement 2.1.258 billing header fingerprint chain and upstream request continuity
committed
sususu
fix(antigravity): tighten replacement offset guard in tool provenance degradation
committed
Luis Pater
fix(auth): preserve concurrent modifications during auth refresh and preparation
committed
Luis Pater
fix(translator/claude): downgrade strict mode when schema misses required properties
committed
Luis Pater
fix(translator/interactions): avoid tool name collisions with Antigravity intrinsic tools
committed
Luis Pater
Merge pull request #5435 from huangruiteng/codex/openai-compat-bounded-rate-limit-waits
committed
Luis Pater
Merge pull request #5452 from deathemperor/docs/add-infinitus
committed
Luis Pater
feat(plugin): omit stream chunk history on payload chunks for schema v5
committed
Luis Pater
fix(translator/claude): handle incomplete status and terminal state on max_tokens
committed
rome-xi
perf(antigravity): batch replay degradation rewrites (#5461)
committed
Luis Pater
fix(translator/claude): defer message_delta and cache streaming usage
committed
Luis Pater
fix(translator/gemini): nest image parts inside functionResponse
committed
Luis Pater
feat(openai): send periodic ping control frames during responses websocket streaming
committed
Luis Pater
feat(codex): support orphan delegation compatibility via orphan-delegation-compatibility
committed
Supra4E8C
feat: add Swiftproxy Sponser
committed
Supra4E8C
feat: 添加 AxisNow 赞助信息及相关图像到 README 文件
committed
Supra4E8C
chore: remove sponsorship images for Claude API and Code0 from README files
committed
sususu98
feat(session): derive distinct branch session ID, parent lineage on Merkle LCP forks, and enhance Codex fork/subagent affinity (#5418) (#5454)
committed
Luis Pater
fix(auth): prevent individual model quota cooldowns from blocking credential
committed
deathemperor
docs: add Infinitus to the projects based on CLIProxyAPI
committed
Luis Pater
feat(routing): make subagent session affinity configurable via session-affinity-subagents
committed
Luis Pater
Merge pull request #5436 from router-for-me/models
committed
hkfires
fix(codex): clear unsupported reasoning levels
committed
hkfires
refactor(kimi): use request thinking helper
committed
hkfires
feat(auth): use home model capabilities for thinking
committed
hkfires
fix(models): preserve home model capability metadata
committed
huangruiteng
fix(openai-compat): honor bounded rate-limit waits
committed
sususu
fix(claude): upgrade default Claude Code baseline and fingerprint to 2.1.258
committed
Luis Pater
fix(auth): enforce minimum cooldown floor and track attempted credentials on 429
committed
Luis Pater
fix(home): map user_credits_insufficient to 402 and user_period_limit_exceeded to 429
committed
Luis Pater
fix(auth): validate access token expiration and retain valid credentials on refresh failure
committed
Luis Pater
chore(registry): remove "minimal" level from dynamic_allowed definitions in models
committed
Luis Pater
feat(registry): add claude fable 5.1 and gemini 3.8 flash models
committed
Luis Pater
fix(antigravity): bypass quota cooldowns and credit hints when cooling is disabled
committed
Luis Pater
fix(antigravity): strip tool config, labels, and session id in token counting
committed
Luis Pater
fix: harden concurrent session handling, listener lifecycle, and token accounting
committed
Luis Pater
fix(antigravity): preserve unsigned gemini thinking blocks with trailing carriers
committed
Luis Pater
fix(antigravity): retry model fetching per endpoint and prioritize daily base url
committed
Luis Pater
fix(antigravity): support native thinking signatures without prefixes in claude translator
committed
Luis Pater
feat(translator): support cache write tokens in claude responses
committed