Mohammad Hijjawi a42752ee38 fix(vlm): treat gpt-6 and later OpenAI models as reasoning models (#5397) (#5398) ...

The OpenAI backend picked max_completion_tokens and dropped temperature
only for the gpt-5/o1/o3/o4 prefixes, so gpt-6-* models were sent
max_tokens and temperature and every session commit failed with a 400
unsupported_parameter error. Match gpt-5 and any later gpt-N generation
instead of listing a single family.

Fixes #5397

2026-09-28 14:51:56 +08:00

..

config

feat(pathlock): adjust pathlock config (#3854)

2026-08-07 13:21:30 +08:00

crypto

fix(crypto): 禁止重写 Vault root key on 临时 (503) read failures (#3423)

2026-07-24 17:30:34 +08:00

embedder

feat(embedding): surface non-symmetric embedding config for VikingDB provider (#1110)

2026-04-16 17:33:21 +08:00

isolated

Feat/自进化（经验记忆）框架重构 (#2503)

2026-07-03 11:27:17 +08:00

models/rerank

retrieval: add Jev rerank provider (#5247)

2026-09-21 19:05:52 +08:00

retrieve

fix(observer): keep empty retrievals diagnostic-only (#3985)

2026-08-13 21:46:31 +08:00

service

chore: split session.py into session sub-modules (#5363)

2026-09-24 20:49:03 +08:00

session

chore: split session.py into session sub-modules (#5363)

2026-09-24 20:49:03 +08:00

session_train

Feat/自进化（经验记忆）框架重构 (#2503)

2026-07-03 11:27:17 +08:00

stats

chore(session): remove session/used reporting API (#5252)

2026-09-21 19:18:18 +08:00

usage_reporter

fix(agent-evolution): recognize namespaced MCP reads (#4349)

2026-08-26 16:13:45 +08:00

__init__.py

lisence: change the main lisence to AGPL-3.0 (#1085)

2026-03-30 14:37:42 +08:00

test_accessors_base.py

reorg: split parser layer to 2-layer: accessor and parser, so that we can reuse more code (#1428)

2026-04-15 10:33:12 +08:00

test_accessors_git.py

feat(config): add runtime configuration system (#5140)

2026-09-20 19:59:16 +08:00

test_accessors_http.py

fix(api): 区分参数错误、锁冲突和内部故障 (#5097)

2026-09-16 16:30:09 +08:00

test_accessors_registry.py

refactor: remove unused and deprecated APIs (#3272)

2026-07-16 10:49:56 +08:00

test_accessors_web_feed.py

feat(resources): ingest whole sites via sitemap / RSS / Atom (#2858)！

2026-06-26 21:13:03 +08:00

test_aml_run_public.py

feat(benchmark): add AML adapter and public evaluation workflows (#5276)

2026-09-22 15:02:25 +08:00

test_aml_server.py

feat(benchmark): add AML adapter and public evaluation workflows (#5276)

2026-09-22 15:02:25 +08:00

test_async_client_cache.py

feat(langchain): add native async integration support (#3536)

2026-07-28 14:35:02 +08:00

test_body_dump_middleware.py

fix(core): harden privacy and background failure handling (#3548)

2026-08-03 16:07:35 +08:00

test_circuit_breaker.py

feat: implement multi-credential priority call design (#2468)

2026-06-15 15:37:03 +08:00

test_cli_uri_windows_safe.py

fix(cli): make resource URI segments Windows-safe like memory uris (#4517 precedent) (#4549)

2026-09-02 11:21:39 +08:00

test_codex_vlm.py

fix(codex): replace retired default model with GPT-5.6 Terra (#5120)

2026-09-17 11:57:16 +08:00

test_cohere_embedder.py

refactor(embedding): remove unused batch APIs (#3260)

2026-07-15 16:44:57 +08:00

test_cohere_rerank.py

fix: ov status shows embedding and rerank models usage (#1191)

2026-04-03 09:00:38 +08:00

test_connector_config.py

Feat/watch task external auth main (#5310)

2026-09-23 14:33:57 +08:00

test_content_visibility.py

fix(core): harden privacy and background failure handling (#3548)

2026-08-03 16:07:35 +08:00

test_context.py

perf(resources): optimize incremental resource ingestion (#5175)

2026-09-21 14:00:51 +08:00

test_create_agfs_client_git.py

feat: 实现 commit、restore、show 文件系统多版本管理功能 (#2756)

2026-06-25 11:28:04 +08:00

test_dashscope_embedder.py

fix(sdk): consolidate model, client, and evaluation correctness (#3550)

2026-07-28 18:11:30 +08:00

test_directory_initializer.py

fix(storage): 启动时初始化完整 agent 目录 (#4956)

2026-09-14 11:46:07 +08:00

test_embedding_config_gemini.py

fix(embedding): report missing gemini dependency (#4402)

2026-08-27 19:52:36 +08:00

test_embedding_config_voyage.py

lisence: change the main lisence to AGPL-3.0 (#1085)

2026-03-30 14:37:42 +08:00

test_embedding_credential_model_override.py

feat: implement multi-credential priority call design (#2468)

2026-06-15 15:37:03 +08:00

test_embedding_vectorize_strategy.py

fix(embedding): deprecate summary_only without dropping content fallback (#5056)

2026-09-16 14:42:21 +08:00

test_encoding_format_embedding.py

feat(embedder): expose encoding_format for OpenAI/Azure providers (#2092)

2026-05-18 14:13:03 +08:00

test_extra_body_embedding.py

feat(embedder): support extra_body passthrough in OpenAI embedder config (#3495)

2026-07-24 15:07:44 +08:00

test_extra_headers_embedding.py

fix(embedding): disable OpenAI SDK retries (#5166)

2026-09-18 16:36:33 +08:00

test_extra_headers_vlm.py

fix(vlm): 透传工作记忆生成错误并复用共享重试 (#5364)

2026-09-24 18:16:46 +08:00

test_failover_embedder_permanent_advance.py

refactor(embedding): remove unused batch APIs (#3260)

2026-07-15 16:44:57 +08:00

test_find_skills_entrypoint.py

fix(retrieve): return one entry per Skill package from context assembly (#5255)

2026-09-22 15:11:23 +08:00

test_fs_service_git_forwarders.py

feat(snapshot): add path diff API, SDK, and CLI (#3451)

2026-07-23 17:45:36 +08:00

test_gemini_embedder.py

refactor(embedding): remove unused batch APIs (#3260)

2026-07-15 16:44:57 +08:00

test_git_config.py

feat: 实现 commit、restore、show 文件系统多版本管理功能 (#2756)

2026-06-25 11:28:04 +08:00

test_image_search.py

fix(media): downsample oversized image model inputs (#3965)

2026-08-12 22:24:57 +08:00

test_jina_embedder.py

refactor(embedding): remove unused batch APIs (#3260)

2026-07-15 16:44:57 +08:00

test_kimi_glm_vlm.py

refactor: remove unused and deprecated APIs (#3272)

2026-07-16 10:49:56 +08:00

test_langchain_async_integration.py

feat(sdk): sync go/ts/python SDKs with server find/search, recall, an… (#3737)

2026-08-24 14:09:11 +08:00

test_langchain_integration.py

feat(sdk): sync go/ts/python SDKs with server find/search, recall, an… (#3737)

2026-08-24 14:09:11 +08:00

test_langchain_lifecycle_resources.py

refactor(langchain): extract standalone integration package (#3685)

2026-08-03 15:01:13 +08:00

test_langchain_package_boundary.py

feat(uri)!: remove uid-less current-user shorthand in favor of viking://~ (#4196)

2026-08-21 19:00:19 +08:00

test_langchain_recording.py

feat(uri)!: remove uid-less current-user shorthand in favor of viking://~ (#4196)

2026-08-21 19:00:19 +08:00

test_langchain_runtime_actor_peer.py

refactor(client): remove Python embedded mode (#3712)

2026-08-10 18:00:00 +08:00

test_litellm_embedder.py

fix(embedding): truncate LiteLLM async vectors (#4146)

2026-08-20 16:37:26 +08:00

test_litellm_vlm_gemini_cache.py

session: add restricted Python DSL extraction protocol and make it the default (#4581)

2026-09-07 20:49:02 +08:00

test_litellm_vlm_provider_detection.py

fix(vlm): preserve explicit Ollama chat routes (#4898)

2026-09-15 16:31:53 +08:00

test_local_embedder.py

refactor(embedding): remove unused batch APIs (#3260)

2026-07-15 16:44:57 +08:00

test_locomo_peer_wiring.py

fix(benchmark): commit_session takes the telemetry keyword, not options (#4498)

2026-09-21 12:01:54 +08:00

test_locomo_progress_utils.py

Feat/自进化（经验记忆）框架重构 (#2503)

2026-07-03 11:27:17 +08:00

test_ls_normalize_virtual_dirs.py

fix(fs): normalize S3 virtual-directory ls rows without modTime (#4875)

2026-09-10 16:36:25 +08:00

test_mcp_tool_annotations.py

feat(mcp): advertise tool behavior annotations (#5075)

2026-09-25 15:32:24 +08:00

test_mcp_tool_error_results.py

fix(mcp): mark whole-call failures as errors (#5078)

2026-09-25 17:08:58 +08:00

test_message_format.py

session: add restricted Python DSL extraction protocol and make it the default (#4581)

2026-09-07 20:49:02 +08:00

test_message.py

refactor: remove unused and deprecated APIs (#3272)

2026-07-16 10:49:56 +08:00

test_mineru_preflight.py

feat(pdf): refactor MinerU parsing to the official file_parse API (#3953)

2026-08-17 13:48:49 +08:00

test_minimax_embedder_simple.py

lisence: change the main lisence to AGPL-3.0 (#1085)

2026-03-30 14:37:42 +08:00

test_model_retry.py

fix(vlm): 透传工作记忆生成错误并复用共享重试 (#5364)

2026-09-24 18:16:46 +08:00

test_namespace_uri_classification.py

fix(namespace): 统一 agent 公共目录并删除旧身份逻辑 (#5069)

2026-09-16 11:56:02 +08:00

test_ollama_embedding_factory.py

lisence: change the main lisence to AGPL-3.0 (#1085)

2026-03-30 14:37:42 +08:00

test_ollama_utils.py

feat(cli): add query planner setup to init wizard (#2485)

2026-06-08 17:12:05 +08:00

test_openai_embedder_http_pool.py

fix(embedder): bound OpenAI HTTP connection pools (#4475)

2026-09-03 16:16:26 +08:00

test_openai_embedder.py

fix: enable explicit openai multimodal embeddings (#4433)

2026-09-04 17:10:32 +08:00

test_ordered_credential_switcher.py

refactor: remove unused and deprecated APIs (#3272)

2026-07-16 10:49:56 +08:00

test_path_safety.py

feat: implement Skill-driven knowledge compilation for ov compile (#3567)

2026-07-28 20:33:25 +08:00

test_path_variables.py

refactor: remove dead code paths (#2545)

2026-06-10 16:34:50 +08:00

test_personamem_evaluate.py

feat(benchmark): add AML adapter and public evaluation workflows (#5276)

2026-09-22 15:02:25 +08:00

test_personamem_prepare.py

feat(benchmark): add AML adapter and public evaluation workflows (#5276)

2026-09-22 15:02:25 +08:00

test_process_lock.py

fix(storage): 用文件锁替代本地存储 PID 检查 (#5159)

2026-09-18 15:10:13 +08:00

test_remote_rollout_executor.py

优化Train相关的逻辑 (#3051)

2026-07-07 11:29:03 +08:00

test_resource_reservation.py

feat(acl): 增加资源 ACL、用户组授权与向量权限过滤 (#3213)

2026-08-28 17:22:55 +08:00

test_search_filter_only_query.py

feat(search): allow find without a query when a filter is given (#4694)

2026-09-04 17:02:52 +08:00

test_search_filters.py

refactor: remove dead code paths (#2545)

2026-06-10 16:34:50 +08:00

test_search_tags_filter.py

fix(tags): remove charset and length restrictions (#4761)

2026-09-07 14:31:05 +08:00

test_server_bootstrap_bot_gateway.py

feat(bot): manage Docker-backed OpenSandbox lifecycle and persistent workspaces (#5269)

2026-09-22 11:37:27 +08:00

test_skill_privacy_cancellation.py

feat(skills): index full skill packages and return the best hit per skill (#5045)

2026-09-18 15:40:55 +08:00

test_skill_processor_nested_skill_md.py

fix(skill): preserve nested SKILL.md files during skill import (#4055)

2026-08-18 10:38:41 +08:00

test_skill_processor_none.py

feat(skills): index full skill packages and return the best hit per skill (#5045)

2026-09-18 15:40:55 +08:00

test_skill_processor_windows_paths.py

fix(session): preserve outer lease when creating extracted skills (#5018)

2026-09-17 18:27:15 +08:00

test_skill_search_contract.py

feat(skills): index full skill packages and return the best hit per skill (#5045)

2026-09-18 15:40:55 +08:00

test_stream_config_vlm.py

fix(vlm): reject unsupported stream configuration (#4541)

2026-08-31 20:52:32 +08:00

test_streaming_batcher.py

fix(storage): preserve deletion and session durability (#3553)

2026-07-31 21:45:29 +08:00

test_summarizer_context_type.py

feat(memory): 引入 User/Peer 记忆隔离模型 (#2236)

2026-06-05 10:55:48 +08:00

test_summarizer_resources_root_split.py

feat(acl): 增加资源 ACL、用户组授权与向量权限过滤 (#3213)

2026-08-28 17:22:55 +08:00

test_time_utils_comprehensive.py

refactor: remove unused and deprecated APIs (#3272)

2026-07-16 10:49:56 +08:00

test_time_utils.py

fix: improve ISO datetime parsing (#404)

2026-03-04 11:30:50 +08:00

test_uri_short_format.py

feat(uri): add viking://~ home alias for the caller's user root (#4167)

2026-08-20 20:44:40 +08:00

test_uri_validation.py

fix(namespace): 统一 agent 公共目录并删除旧身份逻辑 (#5069)

2026-09-16 11:56:02 +08:00

test_vectorize_file_strategy.py

feat(config): isolate account runtime resources (#5323)

2026-09-24 15:08:56 +08:00

test_vikingbot_chat_images.py

feat(bot): support image inputs in VikingBot Chat API (#3619)

2026-07-30 14:00:48 +08:00

test_vikingbot_vlm_adapter_retry.py

fix: improve Ark VLM error diagnostics (#4072)

2026-08-18 13:58:52 +08:00

test_vlm_failover.py

feat(config): isolate account runtime resources (#5323)

2026-09-24 15:08:56 +08:00

test_vlm_reasoning_models.py

fix(vlm): treat gpt-6 and later OpenAI models as reasoning models (#5397) (#5398)

2026-09-28 14:51:56 +08:00

test_vlm_response_formats.py

session: add restricted Python DSL extraction protocol and make it the default (#4581)

2026-09-07 20:49:02 +08:00

test_vlm_thinking_param.py

fix(bot): Tool-call ordering, thinking mode, and image uploads (#2944)

2026-07-02 15:28:42 +08:00

test_volcengine_embedder_input.py

Feat/memory extraction operation metrics (#4410)

2026-08-28 11:44:43 +08:00

test_voyage_embedder.py

refactor(embedding): remove unused batch APIs (#3260)

2026-07-15 16:44:57 +08:00

test_web_importer.py

Fix/add resource (#4344)

2026-08-27 20:58:29 +08:00