Peter Steinberger
1380dc170e
fix(browser): avoid restart hint for external profiles
2026-04-25 19:18:06 +01:00
Vincent Koc
d6ef1fcf24
feat(diagnostics-otel): export tool loop events
2026-04-25 11:11:56 -07:00
Chris Zhang
c3bfd328ad
feat(litellm): add image generation provider ( #70246 )
...
* feat(litellm): add image generation provider
Registers litellm as an image-generation provider so model refs like
litellm/gpt-image-2 route through the LiteLLM proxy, and
agents.defaults.imageGenerationModel.fallbacks entries of the form
litellm/... resolve without "No image-generation provider registered
for litellm" errors.
Implementation uses the OpenAI-compatible /images/generations and
/images/edits endpoints that LiteLLM proxies for. BaseUrl resolves from
models.providers.litellm.baseUrl (default http://localhost:4000 ). Private
network is auto-allowed when baseUrl is a loopback/RFC1918 address, which
covers the common self-hosted LiteLLM proxy case without needing
OPENCLAW_PROVIDER_ALLOW_PRIVATE_NETWORK. Public baseUrls keep normal SSRF
defaults.
Default model is gpt-image-2 (matching upstream 4.21+ OpenAI default).
Advertises the same 2K/4K sizes OpenAI now exposes, plus legacy
256/512/1024 for dall-e-3. Supports both generate and edit.
Local patch. LiteLLM has no upstream image-generation support yet; revisit
if upstream adds one.
* ci: rerun after upstream main hot-fix
* fix(litellm): harden image generation provider
---------
Co-authored-by: Chris Zhang <chris@ChrisdeMac-mini.local >
Co-authored-by: Peter Steinberger <steipete@gmail.com >
2026-04-25 19:06:51 +01:00
Vincent Koc
ff172f46a5
feat(diagnostics-otel): add context assembly spans
2026-04-25 11:03:46 -07:00
Peter Steinberger
afd6b5d6fc
fix(opencode-go): route DeepSeek V4 through OpenAI transport
2026-04-25 18:58:08 +01:00
Peter Steinberger
9ffe764416
fix(whatsapp): send voice note text separately
2026-04-25 18:55:03 +01:00
Peter Steinberger
617e1dd6bf
fix(browser): honor remote CDP open timeouts
2026-04-25 18:52:57 +01:00
Vincent Koc
44114328b4
feat(diagnostics): surface provider request id hashes
2026-04-25 10:46:10 -07:00
Vincent Koc
a7de722f4f
fix(diagnostics-otel): align GenAI semconv attrs
2026-04-25 10:33:13 -07:00
Peter Steinberger
88df8fe09d
fix(browser): clarify Browserless CDP attach handling
2026-04-25 18:26:57 +01:00
Vincent Koc
dcdf97685b
fix(diagnostics): trust internal trace parents ( #71574 )
...
* fix(diagnostics): trust internal trace parents
* fix(diagnostics): harden trusted trace metadata
* fix(tooling): honor explicit oxlint threads
* fix(agents): use stable nonmutating sort helpers
* chore(plugin-sdk): refresh api baseline
* fix(diagnostics): gate internal event subscriptions
* fix(diagnostics): isolate listener event copies
* chore(plugin-sdk): refresh internal diagnostics baseline
* chore(plugin-sdk): refresh diagnostics event baseline
* fix(diagnostics): keep event state module local
* fix(diagnostics): harden internal subscription capability
* fix(diagnostics): freeze listener metadata
2026-04-25 10:18:52 -07:00
Peter Steinberger
67506ac2a9
fix(xai): support video reference images
2026-04-25 18:14:51 +01:00
Peter Steinberger
390be8138f
fix: add OpenCode Go DeepSeek V4 models
2026-04-25 18:11:59 +01:00
Peter Steinberger
6b3e4b88d6
test: update QA parity fixtures for GPT-5.5
2026-04-25 18:05:28 +01:00
Peter Steinberger
8bead989da
fix(telegram): frame audio transcripts as untrusted
2026-04-25 17:45:40 +01:00
Vincent Koc
ab1d1a5c9e
fix(browser): configure Chrome MCP existing-session launch ( #71560 )
2026-04-25 05:46:39 -07:00
Peter Steinberger
dd78b7f773
fix: harden OpenCode ACP bind dispatch
2026-04-25 13:38:58 +01:00
skylee-01
f7b71abf48
fix(agents): pass Claude system prompt via file
2026-04-25 17:59:25 +05:30
Peter Steinberger
b26367e22f
test: add Crestodian QA lab setup scenario
2026-04-25 13:15:11 +01:00
Sahil Satralkar
3064ea78ab
fix(telegram): recover incomplete preview finalization ( #71554 )
...
Fix Telegram partial-stream preview finalization so ambiguous final edit failures fall back to a final send when the visible preview is a strict prefix of the answer.
Includes archived-preview regression coverage and generated config metadata refresh.
Thanks @sahilsatralkar.
Co-authored-by: Sahil Satralkar <62758655+sahilsatralkar@users.noreply.github.com >
2026-04-25 13:01:10 +01:00
Peter Steinberger
e25b3c6056
fix(browser): align bare ws cdp readiness
2026-04-25 13:00:22 +01:00
Peter Steinberger
a48998d8c8
test(qqbot): cover voice utility contracts
2026-04-25 12:57:23 +01:00
Peter Steinberger
c307700db0
test(whatsapp): cover group generated media delivery
2026-04-25 12:56:53 +01:00
Peter Steinberger
6eec38ad5a
feat(discord): allow voice model override
2026-04-25 12:47:46 +01:00
Peter Steinberger
3507efa4ec
fix(media): preserve oversized video generation delivery
2026-04-25 12:41:43 +01:00
Vincent Koc
e174d96cc0
refactor(media): move sharp image ops into media runtime ( #71519 )
...
* refactor(media): move sharp image ops into plugin
* fix(media): pass image pixel budget to sharp plugin
* refactor(media): reuse media understanding sharp runtime
* test(build): allow staged runtime core graphs
2026-04-25 04:31:10 -07:00
Peter Steinberger
b2b898c2a8
feat(browser): configure local startup timeouts
2026-04-25 12:30:35 +01:00
Peter Steinberger
c5fe80ad58
fix: make qa config apply retries idempotent
2026-04-25 12:30:07 +01:00
Ayaan Zaidi
bdba4fa1bf
fix: isolate active memory auth health ( #71539 )
...
* fix(agents): scope helper auth failures
* fix(active-memory): isolate recall auth health
* fix: isolate active memory auth health (#71539 )
* fix: avoid auth policy import cycle (#71539 )
2026-04-25 16:50:38 +05:30
Peter Steinberger
b511250e5c
feat(media): add voice conversion and speech plugins
2026-04-25 12:12:33 +01:00
Vincent Koc
84a22a64be
fix(feishu): finish streaming card closeout
2026-04-25 04:04:03 -07:00
Peter Steinberger
935cd34e9f
fix(openai): omit Azure image deployment model body
2026-04-25 12:02:26 +01:00
Peter Steinberger
89755d1c79
refactor(browser): simplify lazy CLI placeholders
2026-04-25 11:48:59 +01:00
Peter Steinberger
c52ec520c7
feat(browser): add one-shot headless start override
2026-04-25 11:42:03 +01:00
Peter Steinberger
776d2ab65d
fix(browser): lazy-load browser CLI runtime
...
Co-authored-by: pandego <7780875+pandego@users.noreply.github.com >
Co-authored-by: Tianworld <3580442280@qq.com >
2026-04-25 11:40:20 +01:00
Peter Steinberger
a0c70c4f5a
fix(google): guard veo rest polling
2026-04-25 11:17:23 +01:00
Peter Steinberger
9b48e4c0b6
fix(browser): fall back to headless on Linux without display
2026-04-25 11:13:42 +01:00
Peter Steinberger
b5a1b7d44d
fix(google): guard veo video downloads
2026-04-25 11:12:49 +01:00
Peter Steinberger
978f869fcd
fix(google): type veo fallback operation state
2026-04-25 11:11:14 +01:00
Peter Steinberger
94686c63fb
fix(google): fall back to rest for veo sdk 404
2026-04-25 11:11:14 +01:00
Peter Steinberger
5e0cca5e24
fix(google): narrow veo api key for uri download
2026-04-25 11:07:16 +01:00
Peter Steinberger
c11337149b
fix(google): download direct veo video uri
2026-04-25 11:07:16 +01:00
Vincent Koc
455eba7f94
fix(feishu): coalesce streaming card final delivery
2026-04-25 03:06:38 -07:00
Peter Steinberger
38703ed9a1
fix(discord): identify voice attachment metadata
2026-04-25 11:05:38 +01:00
Peter Steinberger
b9ea631b4b
fix(openai): use gpt 5.5 for codex image responses
2026-04-25 11:03:53 +01:00
Peter Steinberger
9fe35a0c62
fix(discord): restore voice note audio preflight
2026-04-25 10:57:37 +01:00
Peter Steinberger
e6713c0a61
test(minimax): cover default music model normalization
2026-04-25 10:55:54 +01:00
Peter Steinberger
ed8384d32d
fix(minimax): default music generation to music 2.6
2026-04-25 10:55:54 +01:00
Yunsu
9c64a0ca23
fix(google): avoid doubled media generation API version
...
Strip configured trailing /v1beta from Google music/video generation base URLs before calling the Google GenAI SDK.\n\nFixes #63240.\n\nThanks @Hybirdss.
2026-04-25 10:45:38 +01:00
Peter Steinberger
a7604f8170
fix(minimax): support token plan tts auth
2026-04-25 10:36:12 +01:00
