opensource/dify
2
0
Fork 0
mirror of https://github.com/langgenius/dify.git synced 2026-05-11 23:18:39 +08:00
Code Issues Packages Projects Releases Wiki Activity
10,150 Commits 494 Branches 170 Tags 710 MiB
feat/new-agent-node
Commit Graph

121 Commits

Author SHA1 Message Date
Yansong Zhang
e9ee897973 fix: resolve remaining CI failures for style checks and unit tests 
- Add model_features property and build_execution_context method to
  AgentAppRunner to fix mypy attr-defined errors
- Export WorkflowComment, WorkflowCommentReply, WorkflowCommentMention
  from models/__init__.py to fix import errors
- Add NestedNodeGraphRequest, NestedNodeGraphResponse,
  NestedNodeParameterSchema to services/workflow/entities.py
- Update test_agent_chat_app_runner: tests for invalid LLM mode and
  invalid strategy now reflect unified AgentAppRunner behavior
  (no longer raises ValueError for these cases)

Made-with: Cursor
 2026-04-13 16:07:38 +08:00
Yansong Zhang
c03b25a940 merge: resolve conflicts with origin/main 
Conflicts resolved:
- workflow_app_runner.py: adopt main's DifyGraphInitContext pattern
- token_buffer_memory.py: adopt main's match/case, add AppMode.AGENT
- app_dsl_service.py: adopt main's match/case, add AppMode.AGENT

Made-with: Cursor
 2026-04-13 12:52:56 +08:00
Yansong Zhang
90cce7693f revert: remove all sandbox and skill related code 
Remove ~12,900 lines of sandbox/skill code that was ported from
feat/support-agent-sandbox. This reverts to direct tool execution
(the original behavior before sandbox integration).

Removed:
- core/sandbox/ (SandboxBuilder, bash tools, providers, initializers)
- core/skill/ (SkillManager, assembler, entities)
- core/virtual_environment/ (5 provider implementations)
- core/zip_sandbox/ (archive operations)
- core/app_assets/ (asset management)
- core/app_bundle/ (bundle management)
- controllers/cli_api/ (DifyCli callback endpoints)
- services/sandbox/ (provider service)
- services/skill_service, app_asset_service, app_bundle_service
- models/sandbox.py, app_asset.py
- bin/dify-cli-* (3 platform binaries)
- web sandbox-provider-page and service
- SandboxLayer, _resolve_sandbox_context, _invoke_tool_in_sandbox
- CliApiConfig, DIFY_SANDBOX_CONTEXT_KEY
- sandbox-related migrations

Preserved: All Agent V2 core functionality (agent-v2 node, strategy
engine, transparent upgrade, LLM remapping, memory, context, tools
via direct execution).

Made-with: Cursor
 2026-04-13 10:42:36 +08:00
wdeveloper16
602753b68a
refactor: replace bare dict with ActionDict TypedDict in cot_agent_runner (#34997) 2026-04-13 00:53:20 +00:00
Yansong Zhang
5cdae671d5 feat(api): integrate Sandbox Provider into Agent V2 execution pipeline 
Close 3 integration gaps between the ported Sandbox system and Agent V2:

1. Fix _invoke_tool_in_sandbox to use SandboxBashSession context manager
   API correctly (keyword args, bash_tool, ToolReference), with graceful
   fallback to direct invocation when DifyCli binary is unavailable.

2. Inject sandbox into run_context via _resolve_sandbox_context() in
   WorkflowBasedAppRunner — automatically creates a sandbox when a
   tenant has an active sandbox provider configured.

3. Register SandboxLayer in both advanced_chat and workflow app runners
   for proper sandbox lifecycle cleanup on graph end.

Also: make SkillInitializer non-fatal when no skill bundle exists,
add node_id to ExecutionContext for sandbox session scoping.

Made-with: Cursor
 2026-04-10 10:14:42 +08:00
Yansong Zhang
3d4be88d97 fix(api): remove unsupported 'user' param from FC/ReAct invoke_llm calls 
FunctionCallStrategy and ReActStrategy were passing user=self.context.user_id
to ModelInstance.invoke_llm() which doesn't accept that parameter.
This caused tool-using agent runs to fail with:
  "ModelInstance.invoke_llm() got an unexpected keyword argument 'user'"

Verified: Agent V2 with current_time tool now works end-to-end:
  ROUND 1: LLM thought -> CALL current_time -> got time
  ROUND 2: LLM generates answer with time info
Made-with: Cursor
 2026-04-09 12:18:07 +08:00
Yansong Zhang
44491e427c feat(api): enable all sandbox/skill controller routes and resolve dependencies (P0) 
Resolve the full dependency chain to enable all previously disabled controllers:

Enabled routes:
- sandbox_files: sandbox file browser API
- sandbox_providers: sandbox provider management API
- app_asset: app asset management API
- skills: skill extraction API
- CLI API blueprint: DifyCli callback endpoints (/cli/api/*)

Dependencies extracted (64 files, ~8000 lines):
- models/sandbox.py, models/app_asset.py: DB models
- core/zip_sandbox/: zip-based sandbox execution
- core/session/: CLI API session management
- core/memory/: base memory + node token buffer
- core/helper/creators.py: helper utilities
- core/llm_generator/: context models, output models, utils
- core/workflow/nodes/command/: command node type
- core/workflow/nodes/file_upload/: file upload node type
- core/app/entities/: app_asset_entities, app_bundle_entities, llm_generation_entities
- services/: asset_content, skill, workflow_collaboration, workflow_comment
- controllers/console/app/error.py: AppAsset error classes
- core/tools/utils/system_encryption.py

Import fixes:
- dify_graph.enums -> graphon.enums in skill_service.py
- get_signed_file_url_for_plugin -> get_signed_file_url in cli_api.py

All 5 controllers verified: import OK, Flask starts successfully.
46 existing tests still pass.

Made-with: Cursor
 2026-04-09 09:36:16 +08:00
Yansong Zhang
96641a93f6 feat(api): add Agent V2 node and new Agent app type (Phase 1-3) 
Introduce a new unified Agent V2 workflow node that combines LLM capabilities
with agent tool-calling loops, along with a new AppMode.AGENT for standalone
agent apps backed by single-node workflows.

Phase 1 — Agent Patterns:
- Add core/agent/patterns/ module (AgentPattern, FunctionCallStrategy,
  ReActStrategy, StrategyFactory) ported from feat/support-agent-sandbox
- Add ExecutionContext, AgentLog, AgentResult entities
- Add Tool.to_prompt_message_tool() for LLM-consumable tool conversion

Phase 2 — Agent V2 Workflow Node:
- Add core/workflow/nodes/agent_v2/ (AgentV2Node, AgentV2NodeData,
  AgentV2ToolManager, AgentV2EventAdapter)
- Register agent-v2 node type in DifyNodeFactory
- No-tools path: single LLM call (LLM Node equivalent)
- Tools path: FC/ReAct loop via StrategyFactory

Phase 3 — Agent App Type:
- Add AppMode.AGENT to model enum
- Add WorkflowGraphFactory for auto-generating start->agent_v2->answer graphs
- AppService.create_app() creates workflow draft for AGENT mode
- AppGenerateService.generate() routes AGENT to AdvancedChatAppGenerator
- Console API and DSL import/export support AGENT mode
- Default app template for AGENT mode

Old agent/agent-chat/LLM node paths are fully preserved.
38 unit tests all passing.

Made-with: Cursor
 2026-04-08 12:31:23 +08:00
Pulakesh
2575a3a3ab
refactor(api): clean up AssistantPromptMessage typing in CotChatAgentRunner (#34681) 2026-04-07 22:53:14 +00:00
Renzo
364d7ebc40
refactor: core/tools, agent, callback_handler, encrypter, llm_generator, plugin, inner_api (#34205) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: Asuka Minato <i@asukaminato.eu.org>
 2026-03-28 10:14:43 +00:00
99
40591a7c50
refactor(api): use standalone graphon package (#34209) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2026-03-27 21:05:32 +00:00
99
52e7492cbc
refactor(api): rename dify_graph to graphon (#34095) 2026-03-25 21:58:56 +08:00
-LAN-
56593f20b0
refactor(api): continue decoupling dify_graph from API concerns (#33580) 
Signed-off-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: WH-2099 <wh2099@pm.me>
 2026-03-25 20:32:24 +08:00
Renzo
7757bb5089
refactor(api): Query API to select function_1 (#33565) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2026-03-17 23:29:16 +09:00
99
1b6e695520
refactor(workflow): move agent node back to core workflow (#33431) 2026-03-14 22:33:13 +08:00
-LAN-
4fd6b52808
refactor(api): move model_runtime into dify_graph (#32858) 2026-03-02 20:15:32 +08:00
-LAN-
c917838f9c
refactor: move workflow package to dify_graph (#32844) 2026-03-02 18:42:30 +08:00
-LAN-
a694533fc9
refactor(workflow): inject credential/model access ports into LLM nodes (#32569) 
Signed-off-by: -LAN- <laipz8200@outlook.com>
 2026-02-27 14:36:41 +08:00
99
7656d514b9
refactor(workflow-file): move core.file to core.workflow.file (#32252) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2026-02-16 22:38:19 +08:00
呆萌闷油瓶
9be863fefa
fix: missing content if assistant message with tool_calls (#30083) 
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2026-01-13 12:46:33 +08:00
-LAN-
1e10bf525c
refactor(models): Refine MessageAgentThought SQLAlchemy typing (#27749) 
Co-authored-by: Asuka Minato <i@asukaminato.eu.org>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2026-01-10 17:17:45 +09:00
wangxiaolei
9b6b2f3195
feat: add AgentMaxIterationError exc (#30423) 2026-01-01 00:40:54 +08:00
QuantumGhost
91667e3c1d
feat(api): Implement EventManager error logging and add coverage (#29204) 
- Ensure `EventManager._notify_layers` logs exceptions instead of silently swallowing them 
  so GraphEngine layer failures surface for debugging
- Introduce unit tests to assert the logger captures the runtime error when collecting events
- Enable the `S110` lint rule to catch `try-except-pass` patterns
- Add proper error logging for existing `try-except-pass` blocks.
 2025-12-08 09:40:40 +08:00
-LAN-
85cda47c70
feat: knowledge pipeline (#25360) 
Signed-off-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: twwu <twwu@dify.ai>
Co-authored-by: crazywoola <100913391+crazywoola@users.noreply.github.com>
Co-authored-by: jyong <718720800@qq.com>
Co-authored-by: Wu Tianwei <30284043+WTW0313@users.noreply.github.com>
Co-authored-by: QuantumGhost <obelisk.reg+git@gmail.com>
Co-authored-by: lyzno1 <yuanyouhuilyz@gmail.com>
Co-authored-by: quicksand <quicksandzn@gmail.com>
Co-authored-by: Jyong <76649700+JohnJyong@users.noreply.github.com>
Co-authored-by: lyzno1 <92089059+lyzno1@users.noreply.github.com>
Co-authored-by: zxhlyh <jasonapring2015@outlook.com>
Co-authored-by: Yongtao Huang <yongtaoh2022@gmail.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: Joel <iamjoel007@gmail.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: nite-knite <nkCoding@gmail.com>
Co-authored-by: Hanqing Zhao <sherry9277@gmail.com>
Co-authored-by: gemini-code-assist[bot] <176961590+gemini-code-assist[bot]@users.noreply.github.com>
Co-authored-by: Harry <xh001x@hotmail.com>
 2025-09-18 12:49:10 +08:00
-LAN-
bab4975809
chore: add ast-grep rule to convert Optional[T] to T | None (#25560) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2025-09-15 13:06:33 +08:00
Krito.
a13d7987e0
chore: adopt StrEnum and auto() for some string-typed enums (#25129) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: crazywoola <100913391+crazywoola@users.noreply.github.com>
 2025-09-12 21:14:26 +08:00
-LAN-
08dd3f7b50
Fix basedpyright type errors (#25435) 
Signed-off-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2025-09-10 01:54:26 +08:00
Asuka Minato
a78339a040
remove bare list, dict, Sequence, None, Any (#25058) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
Co-authored-by: -LAN- <laipz8200@outlook.com>
 2025-09-06 03:32:23 +08:00
Yongtao Huang
be3af1e234
Migrate SQLAlchemy from 1.x to 2.0 with automated and manual adjustments (#23224) 
Co-authored-by: Yongtao Huang <99629139+hyongtao-db@users.noreply.github.com>
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2025-09-02 10:30:19 +08:00
Yongtao Huang
1d09708eb7
Fix incorrect exception reference for json.dumps() (#24329) 
Co-authored-by: Yongtao Huang <99629139+hyongtao-db@users.noreply.github.com>
 2025-08-24 00:31:04 +08:00
jiangbo721
805b698c2e
Feat/chat message image first for agent and advanced_chat APP (#23796) 
Co-authored-by: autofix-ci[bot] <114827586+autofix-ci[bot]@users.noreply.github.com>
 2025-08-22 16:42:15 +08:00
Will
67a0751cf3
fix: Improve create_agent_thought and save_agent_thought Logic (#21263) 2025-07-27 11:06:37 +08:00
Asuka Minato
ef51678c73
orm filter -> where (#22801) 
Signed-off-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: Claude <noreply@anthropic.com>
 2025-07-24 00:57:45 +08:00
Maries
a4ef900916
Support OAuth Integration for Plugin Tools (#22550) 
Co-authored-by: zxhlyh <jasonapring2015@outlook.com>
Co-authored-by: Yeuoly <admin@srmxy.cn>
 2025-07-17 17:18:44 +08:00
Chieh Wang
30aa052a57
feat: Add Citations and Attributions to Agent Node (#18558) 
Co-authored-by: oneness0 <2902216407@qq.com>
Co-authored-by: Novice <novice12185727@gmail.com>
 2025-07-16 15:46:15 +08:00
-LAN-
6eb155ae69
feat(api/repo): Allow to config repository implementation (#21458) 
Signed-off-by: -LAN- <laipz8200@outlook.com>
Co-authored-by: QuantumGhost <obelisk.reg+git@gmail.com>
 2025-07-14 14:54:38 +08:00
Novice
535fff62f3
feat: add MCP support (#20716) 
Co-authored-by: QuantumGhost <obelisk.reg+git@gmail.com>
 2025-07-10 14:01:34 +08:00
Novice
b2ab401279
chore: remove agent turn limits (#19930) 2025-05-29 09:51:56 +08:00
Will
0358859467
fix: llm_usage.total_tokens stat (#19177) 2025-05-07 16:42:49 +08:00
Will
bfa652f2d0
fix: metadata filtering condition variable unassigned; fix External K… (#19208) 2025-05-07 14:52:09 +08:00
kingpc
65b7a783fe
fix: metadata filter not work (#19020) 
Co-authored-by: 金鹏程 <jinpengcheng01@corp.netease.com>
Co-authored-by: crazywoola <427733928@qq.com>
 2025-04-30 11:06:03 +08:00
QuantumGhost
bd1bbfee4b
Enhance Code Consistency Across Repository with .editorconfig (#19023) 2025-04-29 18:04:33 +08:00
Yeuoly
abafa68647
refactor: rename plugin manager to plugin client and rename path from manager to impl (#18876) 2025-04-27 14:22:25 +08:00
ZalterCitty
a1158cc946
fix: Update prompt message content types to use Literal and add union type for content (#17136) 
Co-authored-by: 朱庆超 <zhuqingchao@xiaomi.com>
Co-authored-by: crazywoola <427733928@qq.com>
 2025-04-22 16:17:55 +08:00
Junjie.M
b7e8517b31
feat: agent strategy parameter add help information (#18192) 2025-04-16 17:24:09 +08:00
huangzhuo1949
5dd9acbe44
fix: cot agent chinese json bug (#18073) 
Co-authored-by: huangzhuo <huangzhuo1@xiaomi.com>
 2025-04-15 15:36:44 +08:00
Novice
dcdec98c8e
feat: agent node add memory (#15976) 2025-04-03 16:40:58 +08:00
shirukai
6cf258a809
fix: code block syntax cannot be displayed correctly in react mode (#16904) 2025-03-31 16:27:29 +08:00
jiangbo721
34cba83ac4
fix: bug that overwrote the llm model thought process when final_answ… (#17074) 
Co-authored-by: 刘江波 <jiangbo721@163.com>
 2025-03-28 23:41:32 +08:00
Yeuoly
490b6d092e
Fix/plugin race condition (#14253) 2025-02-25 12:20:47 +08:00