New Features

publish oauth client metadata document (#47737)
add quota reset support action (#47313)
route migrated BYOK through provider connections (#47266)
send X-Interaction-Id header with session id (#47215)
improve market share chart (#47115)
add Muse Spark 1.3 and Gemini 3.8 Flash (#46836)
keep migrated teams on the new Console (#46830)
tolerate Anthropic thinking block binding (#46653)
add workspace migration timestamp (#46627)
animate Go usage allowances and bonuses (#46055)
support Azure CLI authentication (#45079)
load supported v2 config in v1 (#45421)
add weekly retention
add retention metrics
add unblock workspace action
add Ox Alpha to usage graph (#43809)
include free tier usage
promote Hy3 usage (#43429)
native OpenAI and Anthropic passthroughs for Cloudflare AI Gateway (#42634)
add GLM 5.3 (#42518)
add Gemini 3.7 Flash (#42390)
query r2 data catalog
add go usage endpoint (#16513)
open project menu on right-click (#41013)
add global locale coverage (#40992)
promote DeepSeek V4 Flash usage (#40988)
add cursor style configuration (#32295)
export session as json from ui (#40781)
add debug layout direction toggle (#40393)
localize hardcoded application copy (#40377)
add minimal RTL support (#40372)
support locale plural rules (#40370)
expand and audit locale translations (#40362)
refine diff viewer (#40285)
add GPT 5.6 Luna content (#39812)
fix composer attachment fades (#38547)
adds fades to file tree (#38549)
fix file tree button bg (#38629)
disallow duplicate attachments (#39464)
always show new session button (#39520)
migrate to solid-sonner (#39519)
update arrow buttons in review tab top bar (#39469)
discover Modal models (#39066)
refine tab states (#39472)

Bug Fixes

cd does not update footer cwd (LAC-3732) (#44)
publish cwd.updated when caller passes instance directory (LAC-3732)
preserve explicit OpenAI service tiers (#47671)
bump Azure SDK to 3.0.93 (#47664)
bump OpenAI SDK to 3.0.88 (#47659)
add GitLab reasoning variants (#47306)
connect enterprise form to Chatwoot (#47488)
openai usage normalization and tier threshold config (#47342)
compare Codex GPT versions by major and minor (#47385)
allow integer GPT versions in Codex model filter (#47384)
publish cwd.updated when caller passes instance directory (LAC-3732)
hide stealth model providers
backport usage reset boundary fix to dev (#47267)
keep omen-alpha under unknown provider (#47248)
use unicode ellipses in interface text (#45126)
reduce database query load
proxy migrated model discovery to v1 endpoint (#47065)
identify desktop in Console device auth (v1) (#47000)
increase open-in icon size (#46540)
default header timeout to five minutes (#46903)
default chunk timeout to five minutes (#46890)
let config opt out of Anthropic thinking blockBinding (#46820)
restore migrated inference proxy requests (#46854)
scope thinking binding to Claude 5.1+ (#46848)
handle SSE reader cancel rejections (#44944)
bump happy-dom to fix GC-dependent MutationObserver flake (#46675)
Fix for incorrect time.start reset in tool call logging (#32574) (#32596)
allow none reasoning effort in Bedrock SDK (#46671)
bump @ai-sdk/amazon-bedrock to 4.0.166 for reasoning and replay fixes (#45520)
stop Azure model discovery from logging to stdout (#46646)
prevent comparison legend collapse
number Go usage requirements
restore documentation list markers
omit empty apply patch move path (#45329)
pin diff highlights query (#46519)
merge deepseek flash variants (#46446)
keep home shortcuts right-aligned (#36906)
backport session rename and tab menu fixes to v1 (#46116)
remove Hy3 Free docs and correct Go chart rendering (#46221)
remove Bun dependency from Azure authentication (#45845)
compare config snapshots as JSON (#45784)
filter unreplayable Bedrock reasoning before caching (#45769)
improve Go comparison chart on mobile (#45044)
map ox alpha to glm 5.3 flash (#45542)
merge duplicate Go usage rows (#45503)
secure server action redirects (#45374)
align retention columns
reduce retention query scan
validate auth redirects (#45027)
recover legacy database migration history (#45061)
rate limit checkout session creation (#45007)
restore focus in stacked dialogs (#44928)
drop archived sessions from home list right away (#44905)
route non-native Cloudflare AI Gateway providers via the REST API (#44828)
support immutable OIDC subjects (#44776)
send Anthropic's dashed native slug through the AI Gateway (#44281)
send parent session header (#44752)
normalize upgrade endpoint (#44686)
discontinue first-month Go discount (#44633)
set duplex for streamed zen requests
support DeepSeek weekend pricing (#44305)
revert scanned zen request streaming (#44470)
release streamed zen prefixes (#44465)
scan zen model before streaming (#44463)
revert streamed zen request bodies (#44444)
stream zen request bodies (#44429)
normalize model casing
reduce zen request memory (#44403)
keep model provider headers visible (#44115)
resolve console device URLs (#44029)
guard textVerbosity injection for @ai-sdk/openai-compatible providers (#43915)
improve chart tooltip truncation
bump Amazon Bedrock provider (#43909)
cache desktop downloads
continue unknown finish responses (#43892)
merge renamed model data (#43883)
route Vertex multi-regions through REP (#42648)
align Ox Alpha model ID (#43837)
preserve file search results while loading (#43836)
register archive session command in both layouts (#41741)
surface subagent tool errors (#43821)
retry raw network finish errors (#43813)
retry network error variants (#43806)
surface resumable subagent errors (#43657)
preserve Cerebras completion limit (#43736)
answer subagent permissions in run (#43675)
align Ox Alpha free model ID (#43695)
clarify market share providers (#43647)
retry xAI capacity stream errors (#43640)
remove encrypted label from thought header (#43588)
show encrypted reasoning status (#43578)
normalize routed model ids (#43582)
remove Qwen sampling defaults (#43310)
restore session request headers (#43188)
properly show authed providers on /connect command (#39915)
ignore malformed model costs (#43248)
preserve inference sessions (#43124)
keep server details editable (#43169)
stop legacy preview publishing
update Google Vertex SDK (#43108)
fall back on oversized websocket requests (#43099)
codex data residency (#42432)
update default console URL (#43043)
correct YouTube footer link (#42941)
enable web search for Go (#42630)
preserve v1 database compatibility (#42444)
remove Gemini 3.7 Flash (#42393)
preserve unicode in grep previews (#42356)
Update global config path in documentation (#42337)
select Kimi prompt by provider (#42161)
pass through reasoning effort (#42166)
pass through reasoning effort (#42164)
pass through reasoning effort (#42160)
adjust instructions and structure to be more clear to smaller models like dsv4 flash (#42045)
authenticate pulumi downloads
correct r2 daily totals
add Merge Gateway reasoning variants (#41867)
cap session retries with jitter (#41939)
detect Copilot PDF input support (#41522)
update blocked account message (#41819)
scope DeepSeek V4 Flash sampling defaults (#41620)
correct OC-2 weak icon color (#41504)
route all Muse family models to the Meta system prompt (#41581)
use widely recognized developer terminology (#41532)
use current default model (#38603)
fall back after full sync failure (#41411)
ignore unknown config fields (#41312)
populate project picker from home (#41158)
use 词元 instead of 令牌 for token in zh locale (#40977)
register new home projects (#41018)
preserve macOS app on window close (#40974)
use tokens for button loading state (#40884)
use chronological message boundaries (#41006)
order stored messages by creation time (#41001)
order messages by creation time (#40994)
sort sessions by persisted time (#41000)
order shared messages by creation time (#40995)
order legacy message loop by time (#40990)
use chronological message boundaries (#40991)
use file times for truncation cleanup (#40987)
keep locale names native (#40985)
complete translation coverage (#40981)
serialize orphaned compaction history (#40800)
update homepage stats
update github stars
support copying over ssh with set-clipboard on tmux config (#30472)
disable packaged console logging (#40794)
embed version in server sidecar (#40764)
allow blob attachments in web UI (#40692)
improve page speed
preserve compatible stream errors (#40718)
reduce html payloads
expand retryable error patterns (#40707)
show pending tool details (#40603)
include cache writes in ACP usage (#40450)
collapse deletion-only edit parts (#40536)
log upstream 5xx bodies from proxied workspace requests (#40135)
don't forward host directory to remote workspace (#40136)
drain updates before end turn (#40422)
correct RTL layout interactions (#40410)
hide zero-usage countries (#40400)
complete shared RTL primitives (#40388)
supply fallback session titles (#40385)
move markdown parsing to worker (#40356)
include pull request identity in context (#40363)
gpt-5.5+ in combination with azure fails with reasoningEffort (#40265)
update OpenRouter logo (#40313)
reuse imported macOS keychain (#40221)
search every known project in the open project dialog (#40202)
clear stale timeline errors (#39465)
skip legacy config reads for v2 (#40211)
prevent stale prompt control reads (#39842)
restore limits graph axis (#39870)
open external links in system browser (#39820)
prevent stale session tab reads (#39767)
open legacy picker at home (#39804)
prevent file tree tab clipping (#39770)
align debug gutter text (#39782)
stop MCP SSE error reconnect loops (#39697)
widen interleaved reasoning fields (#39556)
use OpenCode model catalog URL (#39672)
repair malformed persisted tabs (#39645)
add public fetch compatibility flag
defer model variant selector (#39445)
preserve agent picker for existing users (#39300)

Performance

fix composer lag via buffered blob draft storage (#40207)

Improvements

use OpenTUI Dynamic in session view (#46649)
stream zen bodies without modifiers (#44472)
simplify go usage response
simplify retry error matching (#40694)

Documentation

document client session compatibility
add localized refund FAQs for Go and Zen (#47575)
fix bold rendering by adding spaces around asterisks (#46231)
renew DeepSeek ZDR through September 2026 (#47288)
add Omen Alpha (#47220)
translate Go usage requirements (#47057)
add Eden AI to the providers list (#43386)
add Claude Fable 5.1 (#46728)
update Qwen3.7 Max Go usage (#46555)
add 2026-08-31 sync entry to UPSTREAM_SYNC.md history
end Hy3 usage promotion (#46213)
add Ling 3.0 Flash Fin Free (#45923)
add Hy4 preview (#45904)
add Qwen3.8 Flash (#45836)
add GLM-5.3-Flash (#45269)
remove Ox Alpha Free (#45221)
mention Exa and Parallel as web search backends (#38395)
add Grok 4.6 (#45042)
prohibit abusive multi-account use
clarify prompt data handling (#44854)
correct action token configuration (#44793)
add 2026-08-24 sync entry to UPSTREAM_SYNC.md history
update Zed custom agent config (#44658)
add LongCat-2.0 (#44636)
clarify DeepSeek weekend pricing (#44637)
update GPT 5.6 Sol pricing (#43986)
reflect GPT 5.6 Sol discount (#43874)
add DeepSeek vision model (#43866)
add Ox Alpha Free (#43798)
add Ox Alpha free model (#43690)
end Luna usage promotion (#43351)
add SCX.ai to the providers list (#42520)
add 2026-08-17 sync entry to UPSTREAM_SYNC.md history
update DeepSeek V4 pricing (#42881)
remove DeepSeek Flash promotion (#42858)
add Muse Spark 1.2 (#42508)
sort Gemini 3.7 before 3.6 (#42473)
use responses API for Grok 4.5 (#42373)
remove Ling 3.0 Tiny free model (#42314)
clarify DeepSeek ZDR coverage (#42085)
fix provider display name and PAT typos (#42034)
fix broken DigitalOcean and Daytona links (#42048)
add Grok 4.6
remove expired free models (#41943)
add Hy3 Free (#41814)
add Nemotron 3.5 Lightning (#41750)
add 2026-08-10 sync entry to UPSTREAM_SYNC.md history
replace Ling Flash with Tiny (#41027)
add LongCat free model (#40585)
add RTL development skill (#40543)
add 2026-08-04 sync entry to UPSTREAM_SYNC.md history
add Poolside provider setup (#39187)
remove obsolete vouch system section (#40266)
add Qwen3.8 Max (#40228)
update DeepSeek privacy policy (#40120)
add Modal provider setup (#39710)
add opencode-tavily plugin (#38709)

Maintenance

generate
generate
update nix node_modules hashes
update nix node_modules hashes
generate
generate
generate
generate
generate
generate
update nix node_modules hashes
bump gitlab-ai-provider to 6.13.0 (#46914)
generate
update nix node_modules hashes
update nix node_modules hashes
update nix node_modules hashes
update nix node_modules hashes
use native runtime conditions in development (#46644)
generate
generate
generate
update nix node_modules hashes
generate
generate
add sst unlock workflow
generate
update nix node_modules hashes
update nix node_modules hashes
generate
generate
generate
generate
generate
generate
generate
generate
generate
update nix node_modules hashes
generate
generate
generate
generate
generate
generate
generate
generate
generate
generate
update nix node_modules hashes
generate
remove scheduled beta sync
generate
update nix node_modules hashes
update ai-gateway-provider to 3.2.0 (#42488)
generate
generate
generate
update nix node_modules hashes
add neriousy to team members (#42107)
generate
generate
generate
use DeepSeek for beta conflicts (#41669)
build beta from v2 (#41627)
generate
generate
generate
generate
update nix node_modules hashes
generate
generate
generate
generate
generate
generate
generate
update nix node_modules hashes
generate
update nix node_modules hashes
generate
generate
generate
generate
generate
generate
update nix node_modules hashes
generate
update nix node_modules hashes
generate
update nix node_modules hashes
generate
update nix node_modules hashes
bump gitlab-ai-provider to 6.12.1 (#39490)
generate
generate
update nix node_modules hashes
generate

Testing

disable npm audits in the test preload (#47222)
fix session tools test typecheck and runtime (#46677)
guard patched dependency versions (#46673)
use native config path in permission assertion (#45849)
align unknown finish coverage (#43895)
remove flaky subagent test (#43819)
harden flaky e2e synchronization (#40556)
cover Azure completion reasoning (#40340)

Changes

docs
docs
update zen docs
update zen models
sync release versions for v1.18.29
ci+docs: make lash the working branch — run CI on lash, playbook targets lash, not dev
sync release versions for v1.18.28
sync release versions for v1.18.27
sync release versions for v1.18.26
remove azure discovery stuff (#46666)
docs on proper usage of OpenCode Go
sync release versions for v1.18.25
sync release versions for v1.18.24
fix map inaccuracy
fix map inaccuracy
remove map
update inference headers
update inference headers
update inference headers
sync release versions for v1.18.23
update inference headers
add client header replacement
update model parser
Revert "delay removing first month discount"
sync release versions for v1.18.22
delay removing first month discount
discontinue first month discount
always allow ox alpha in go
sync release versions for v1.18.21
sync release versions for v1.18.20
update muse country list
sync release versions for v1.18.19
sync
sync
doc muse regions
sync
country check for muse spark in free
Revert "update go models"
Revert "chore: generate"
update go models
update zen model name
sync
sync
sync
sync
sync
sync
sync
update ds flash limit
update grok endpoint
sync release versions for v1.18.18
sync release versions for v1.18.17
update ds v4 pro
sync release versions for v1.18.16
sync release versions for v1.18.15
sync
sync
fix log processor
sync release versions for v1.18.14
sync release versions for v1.18.13
sync release versions for v1.18.12
Revert "fix slow queries"
fix slow queries
fix slow queries
fix slow queries
fix gemini reasoning tokens not counted
sync release versions for v1.18.11
deepseek v4 flash
sync release versions for v1.18.10
sync release versions for v1.18.9

New Features

publish oauth client metadata document (#47737)

add quota reset support action (#47313)

route migrated BYOK through provider connections (#47266)

send X-Interaction-Id header with session id (#47215)

improve market share chart (#47115)

add Muse Spark 1.3 and Gemini 3.8 Flash (#46836)

keep migrated teams on the new Console (#46830)

tolerate Anthropic thinking block binding (#46653)

add workspace migration timestamp (#46627)

animate Go usage allowances and bonuses (#46055)

support Azure CLI authentication (#45079)

load supported v2 config in v1 (#45421)

add weekly retention

add retention metrics

add unblock workspace action

add Ox Alpha to usage graph (#43809)

include free tier usage

promote Hy3 usage (#43429)

native OpenAI and Anthropic passthroughs for Cloudflare AI Gateway (#42634)

add GLM 5.3 (#42518)

add Gemini 3.7 Flash (#42390)

query r2 data catalog

add go usage endpoint (#16513)

open project menu on right-click (#41013)

add global locale coverage (#40992)

promote DeepSeek V4 Flash usage (#40988)

add cursor style configuration (#32295)

export session as json from ui (#40781)

add debug layout direction toggle (#40393)

localize hardcoded application copy (#40377)

add minimal RTL support (#40372)

support locale plural rules (#40370)

expand and audit locale translations (#40362)

refine diff viewer (#40285)

add GPT 5.6 Luna content (#39812)

fix composer attachment fades (#38547)

adds fades to file tree (#38549)

fix file tree button bg (#38629)

disallow duplicate attachments (#39464)

always show new session button (#39520)

migrate to solid-sonner (#39519)

update arrow buttons in review tab top bar (#39469)

discover Modal models (#39066)

refine tab states (#39472)

Bug Fixes

cd does not update footer cwd (LAC-3732) (#44)

publish cwd.updated when caller passes instance directory (LAC-3732)

preserve explicit OpenAI service tiers (#47671)

bump Azure SDK to 3.0.93 (#47664)

bump OpenAI SDK to 3.0.88 (#47659)

add GitLab reasoning variants (#47306)

connect enterprise form to Chatwoot (#47488)

openai usage normalization and tier threshold config (#47342)

compare Codex GPT versions by major and minor (#47385)

allow integer GPT versions in Codex model filter (#47384)

publish cwd.updated when caller passes instance directory (LAC-3732)

hide stealth model providers

backport usage reset boundary fix to dev (#47267)

keep omen-alpha under unknown provider (#47248)

use unicode ellipses in interface text (#45126)

reduce database query load

proxy migrated model discovery to v1 endpoint (#47065)

identify desktop in Console device auth (v1) (#47000)

increase open-in icon size (#46540)

default header timeout to five minutes (#46903)

default chunk timeout to five minutes (#46890)

let config opt out of Anthropic thinking blockBinding (#46820)

restore migrated inference proxy requests (#46854)

scope thinking binding to Claude 5.1+ (#46848)

handle SSE reader cancel rejections (#44944)

bump happy-dom to fix GC-dependent MutationObserver flake (#46675)

Fix for incorrect time.start reset in tool call logging (#32574) (#32596)

allow none reasoning effort in Bedrock SDK (#46671)

bump @ai-sdk/amazon-bedrock to 4.0.166 for reasoning and replay fixes (#45520)

stop Azure model discovery from logging to stdout (#46646)

prevent comparison legend collapse

number Go usage requirements

restore documentation list markers

omit empty apply patch move path (#45329)

pin diff highlights query (#46519)

merge deepseek flash variants (#46446)

keep home shortcuts right-aligned (#36906)

backport session rename and tab menu fixes to v1 (#46116)

remove Hy3 Free docs and correct Go chart rendering (#46221)

remove Bun dependency from Azure authentication (#45845)

compare config snapshots as JSON (#45784)

filter unreplayable Bedrock reasoning before caching (#45769)

improve Go comparison chart on mobile (#45044)

map ox alpha to glm 5.3 flash (#45542)

merge duplicate Go usage rows (#45503)

secure server action redirects (#45374)

align retention columns

reduce retention query scan

validate auth redirects (#45027)

recover legacy database migration history (#45061)

rate limit checkout session creation (#45007)

restore focus in stacked dialogs (#44928)

drop archived sessions from home list right away (#44905)

route non-native Cloudflare AI Gateway providers via the REST API (#44828)

support immutable OIDC subjects (#44776)

send Anthropic's dashed native slug through the AI Gateway (#44281)

send parent session header (#44752)

normalize upgrade endpoint (#44686)

discontinue first-month Go discount (#44633)

set duplex for streamed zen requests

support DeepSeek weekend pricing (#44305)

revert scanned zen request streaming (#44470)

release streamed zen prefixes (#44465)

scan zen model before streaming (#44463)

revert streamed zen request bodies (#44444)

stream zen request bodies (#44429)

normalize model casing

reduce zen request memory (#44403)

keep model provider headers visible (#44115)

resolve console device URLs (#44029)

guard textVerbosity injection for @ai-sdk/openai-compatible providers (#43915)

improve chart tooltip truncation

bump Amazon Bedrock provider (#43909)

cache desktop downloads

continue unknown finish responses (#43892)

merge renamed model data (#43883)

route Vertex multi-regions through REP (#42648)

align Ox Alpha model ID (#43837)

preserve file search results while loading (#43836)

surface subagent tool errors (#43821)

retry raw network finish errors (#43813)

retry network error variants (#43806)

surface resumable subagent errors (#43657)

preserve Cerebras completion limit (#43736)

answer subagent permissions in run (#43675)

align Ox Alpha free model ID (#43695)

clarify market share providers (#43647)

retry xAI capacity stream errors (#43640)

remove encrypted label from thought header (#43588)

show encrypted reasoning status (#43578)

normalize routed model ids (#43582)

remove Qwen sampling defaults (#43310)

restore session request headers (#43188)

properly show authed providers on /connect command (#39915)

ignore malformed model costs (#43248)

preserve inference sessions (#43124)

keep server details editable (#43169)

stop legacy preview publishing

update Google Vertex SDK (#43108)

fall back on oversized websocket requests (#43099)

codex data residency (#42432)

update default console URL (#43043)

correct YouTube footer link (#42941)

enable web search for Go (#42630)

preserve v1 database compatibility (#42444)

remove Gemini 3.7 Flash (#42393)

preserve unicode in grep previews (#42356)

Update global config path in documentation (#42337)

select Kimi prompt by provider (#42161)

pass through reasoning effort (#42166)

pass through reasoning effort (#42164)

pass through reasoning effort (#42160)

adjust instructions and structure to be more clear to smaller models like dsv4 flash (#42045)

authenticate pulumi downloads

correct r2 daily totals

add Merge Gateway reasoning variants (#41867)

cap session retries with jitter (#41939)

detect Copilot PDF input support (#41522)

update blocked account message (#41819)

scope DeepSeek V4 Flash sampling defaults (#41620)

correct OC-2 weak icon color (#41504)

route all Muse family models to the Meta system prompt (#41581)

use widely recognized developer terminology (#41532)

use current default model (#38603)

fall back after full sync failure (#41411)

ignore unknown config fields (#41312)

populate project picker from home (#41158)

use 词元 instead of 令牌 for token in zh locale (#40977)

preserve macOS app on window close (#40974)

use tokens for button loading state (#40884)

use chronological message boundaries (#41006)

order stored messages by creation time (#41001)

order messages by creation time (#40994)

sort sessions by persisted time (#41000)

order shared messages by creation time (#40995)

order legacy message loop by time (#40990)

use chronological message boundaries (#40991)

use file times for truncation cleanup (#40987)

keep locale names native (#40985)

complete translation coverage (#40981)

serialize orphaned compaction history (#40800)

update homepage stats

update github stars

support copying over ssh with set-clipboard on tmux config (#30472)

disable packaged console logging (#40794)

embed version in server sidecar (#40764)

allow blob attachments in web UI (#40692)

improve page speed

preserve compatible stream errors (#40718)

reduce html payloads

expand retryable error patterns (#40707)

show pending tool details (#40603)

include cache writes in ACP usage (#40450)

collapse deletion-only edit parts (#40536)

log upstream 5xx bodies from proxied workspace requests (#40135)

don't forward host directory to remote workspace (#40136)

drain updates before end turn (#40422)

correct RTL layout interactions (#40410)

hide zero-usage countries (#40400)

complete shared RTL primitives (#40388)

supply fallback session titles (#40385)

move markdown parsing to worker (#40356)

include pull request identity in context (#40363)

gpt-5.5+ in combination with azure fails with reasoningEffort (#40265)

update OpenRouter logo (#40313)

reuse imported macOS keychain (#40221)

search every known project in the open project dialog (#40202)

clear stale timeline errors (#39465)

skip legacy config reads for v2 (#40211)

prevent stale prompt control reads (#39842)

restore limits graph axis (#39870)

open external links in system browser (#39820)

prevent stale session tab reads (#39767)

open legacy picker at home (#39804)

prevent file tree tab clipping (#39770)

align debug gutter text (#39782)

stop MCP SSE error reconnect loops (#39697)

widen interleaved reasoning fields (#39556)

use OpenCode model catalog URL (#39672)

repair malformed persisted tabs (#39645)

add public fetch compatibility flag

defer model variant selector (#39445)

preserve agent picker for existing users (#39300)

Documentation

document client session compatibility

add localized refund FAQs for Go and Zen (#47575)

fix bold rendering by adding spaces around asterisks (#46231)

renew DeepSeek ZDR through September 2026 (#47288)

add Omen Alpha (#47220)

translate Go usage requirements (#47057)

add Eden AI to the providers list (#43386)

add Claude Fable 5.1 (#46728)

update Qwen3.7 Max Go usage (#46555)

add 2026-08-31 sync entry to UPSTREAM_SYNC.md history

end Hy3 usage promotion (#46213)

add Ling 3.0 Flash Fin Free (#45923)

add Hy4 preview (#45904)

add Qwen3.8 Flash (#45836)

add GLM-5.3-Flash (#45269)

remove Ox Alpha Free (#45221)

mention Exa and Parallel as web search backends (#38395)

add Grok 4.6 (#45042)

prohibit abusive multi-account use

clarify prompt data handling (#44854)

correct action token configuration (#44793)

add 2026-08-24 sync entry to UPSTREAM_SYNC.md history

update Zed custom agent config (#44658)

add LongCat-2.0 (#44636)

clarify DeepSeek weekend pricing (#44637)

update GPT 5.6 Sol pricing (#43986)

reflect GPT 5.6 Sol discount (#43874)

add DeepSeek vision model (#43866)

add Ox Alpha Free (#43798)

add Ox Alpha free model (#43690)

end Luna usage promotion (#43351)

add SCX.ai to the providers list (#42520)

add 2026-08-17 sync entry to UPSTREAM_SYNC.md history

update DeepSeek V4 pricing (#42881)

remove DeepSeek Flash promotion (#42858)

add Muse Spark 1.2 (#42508)

sort Gemini 3.7 before 3.6 (#42473)

use responses API for Grok 4.5 (#42373)

remove Ling 3.0 Tiny free model (#42314)

clarify DeepSeek ZDR coverage (#42085)

fix provider display name and PAT typos (#42034)

fix broken DigitalOcean and Daytona links (#42048)

add Grok 4.6

remove expired free models (#41943)

add Hy3 Free (#41814)

add Nemotron 3.5 Lightning (#41750)

add 2026-08-10 sync entry to UPSTREAM_SYNC.md history

replace Ling Flash with Tiny (#41027)

add LongCat free model (#40585)

add RTL development skill (#40543)

add 2026-08-04 sync entry to UPSTREAM_SYNC.md history

add Poolside provider setup (#39187)

remove obsolete vouch system section (#40266)

add Qwen3.8 Max (#40228)

update DeepSeek privacy policy (#40120)

add Modal provider setup (#39710)

add opencode-tavily plugin (#38709)

Maintenance

generate