Introduce Plugins (#13836)

Signed-off-by: yihong0618 <zouzou0208@gmail.com> Signed-off-by: -LAN- <laipz8200@outlook.com> Signed-off-by: xhe <xw897002528@gmail.com> Signed-off-by: dependabot[bot] <support@github.com> Co-authored-by: takatost <takatost@gmail.com> Co-authored-by: kurokobo <kuro664@gmail.com> Co-authored-by: Novice Lee <novicelee@NoviPro.local> Co-authored-by: zxhlyh <jasonapring2015@outlook.com> Co-authored-by: AkaraChen <akarachen@outlook.com> Co-authored-by: Yi <yxiaoisme@gmail.com> Co-authored-by: Joel <iamjoel007@gmail.com> Co-authored-by: JzoNg <jzongcode@gmail.com> Co-authored-by: twwu <twwu@dify.ai> Co-authored-by: Hiroshi Fujita <fujita-h@users.noreply.github.com> Co-authored-by: AkaraChen <85140972+AkaraChen@users.noreply.github.com> Co-authored-by: NFish <douxc512@gmail.com> Co-authored-by: Wu Tianwei <30284043+WTW0313@users.noreply.github.com> Co-authored-by: 非法操作 <hjlarry@163.com> Co-authored-by: Novice <857526207@qq.com> Co-authored-by: Hiroki Nagai <82458324+nagaihiroki-git@users.noreply.github.com> Co-authored-by: Gen Sato <52241300+halogen22@users.noreply.github.com> Co-authored-by: eux <euxuuu@gmail.com> Co-authored-by: huangzhuo1949 <167434202+huangzhuo1949@users.noreply.github.com> Co-authored-by: huangzhuo <huangzhuo1@xiaomi.com> Co-authored-by: lotsik <lotsik@mail.ru> Co-authored-by: crazywoola <100913391+crazywoola@users.noreply.github.com> Co-authored-by: nite-knite <nkCoding@gmail.com> Co-authored-by: Jyong <76649700+JohnJyong@users.noreply.github.com> Co-authored-by: github-actions[bot] <41898282+github-actions[bot]@users.noreply.github.com> Co-authored-by: gakkiyomi <gakkiyomi@aliyun.com> Co-authored-by: CN-P5 <heibai2006@gmail.com> Co-authored-by: CN-P5 <heibai2006@qq.com> Co-authored-by: Chuehnone <1897025+chuehnone@users.noreply.github.com> Co-authored-by: yihong <zouzou0208@gmail.com> Co-authored-by: Kevin9703 <51311316+Kevin9703@users.noreply.github.com> Co-authored-by: -LAN- <laipz8200@outlook.com> Co-authored-by: Boris Feld <lothiraldan@gmail.com> Co-authored-by: mbo <himabo@gmail.com> Co-authored-by: mabo <mabo@aeyes.ai> Co-authored-by: Warren Chen <warren.chen830@gmail.com> Co-authored-by: JzoNgKVO <27049666+JzoNgKVO@users.noreply.github.com> Co-authored-by: jiandanfeng <chenjh3@wangsu.com> Co-authored-by: zhu-an <70234959+xhdd123321@users.noreply.github.com> Co-authored-by: zhaoqingyu.1075 <zhaoqingyu.1075@bytedance.com> Co-authored-by: 海狸大師 <86974027+yenslife@users.noreply.github.com> Co-authored-by: Xu Song <xusong.vip@gmail.com> Co-authored-by: rayshaw001 <396301947@163.com> Co-authored-by: Ding Jiatong <dingjiatong@gmail.com> Co-authored-by: Bowen Liang <liangbowen@gf.com.cn> Co-authored-by: JasonVV <jasonwangiii@outlook.com> Co-authored-by: le0zh <newlight@qq.com> Co-authored-by: zhuxinliang <zhuxinliang@didiglobal.com> Co-authored-by: k-zaku <zaku99@outlook.jp> Co-authored-by: luckylhb90 <luckylhb90@gmail.com> Co-authored-by: hobo.l <hobo.l@binance.com> Co-authored-by: jiangbo721 <365065261@qq.com> Co-authored-by: 刘江波 <jiangbo721@163.com> Co-authored-by: Shun Miyazawa <34241526+miya@users.noreply.github.com> Co-authored-by: EricPan <30651140+Egfly@users.noreply.github.com> Co-authored-by: crazywoola <427733928@qq.com> Co-authored-by: sino <sino2322@gmail.com> Co-authored-by: Jhvcc <37662342+Jhvcc@users.noreply.github.com> Co-authored-by: lowell <lowell.hu@zkteco.in> Co-authored-by: Boris Polonsky <BorisPolonsky@users.noreply.github.com> Co-authored-by: Ademílson Tonato <ademilsonft@outlook.com> Co-authored-by: Ademílson Tonato <ademilson.tonato@refurbed.com> Co-authored-by: IWAI, Masaharu <iwaim.sub@gmail.com> Co-authored-by: Yueh-Po Peng (Yabi) <94939112+y10ab1@users.noreply.github.com> Co-authored-by: Jason <ggbbddjm@gmail.com> Co-authored-by: Xin Zhang <sjhpzx@gmail.com> Co-authored-by: yjc980121 <3898524+yjc980121@users.noreply.github.com> Co-authored-by: heyszt <36215648+hieheihei@users.noreply.github.com> Co-authored-by: Abdullah AlOsaimi <osaimiacc@gmail.com> Co-authored-by: Abdullah AlOsaimi <189027247+osaimi@users.noreply.github.com> Co-authored-by: Yingchun Lai <laiyingchun@apache.org> Co-authored-by: Hash Brown <hi@xzd.me> Co-authored-by: zuodongxu <192560071+zuodongxu@users.noreply.github.com> Co-authored-by: Masashi Tomooka <tmokmss@users.noreply.github.com> Co-authored-by: aplio <ryo.091219@gmail.com> Co-authored-by: Obada Khalili <54270856+obadakhalili@users.noreply.github.com> Co-authored-by: Nam Vu <zuzoovn@gmail.com> Co-authored-by: Kei YAMAZAKI <1715090+kei-yamazaki@users.noreply.github.com> Co-authored-by: TechnoHouse <13776377+deephbz@users.noreply.github.com> Co-authored-by: Riddhimaan-Senapati <114703025+Riddhimaan-Senapati@users.noreply.github.com> Co-authored-by: MaFee921 <31881301+2284730142@users.noreply.github.com> Co-authored-by: te-chan <t-nakanome@sakura-is.co.jp> Co-authored-by: HQidea <HQidea@users.noreply.github.com> Co-authored-by: Joshbly <36315710+Joshbly@users.noreply.github.com> Co-authored-by: xhe <xw897002528@gmail.com> Co-authored-by: weiwenyan-dev <154779315+weiwenyan-dev@users.noreply.github.com> Co-authored-by: ex_wenyan.wei <ex_wenyan.wei@tcl.com> Co-authored-by: engchina <12236799+engchina@users.noreply.github.com> Co-authored-by: engchina <atjapan2015@gmail.com> Co-authored-by: dependabot[bot] <49699333+dependabot[bot]@users.noreply.github.com> Co-authored-by: 呆萌闷油瓶 <253605712@qq.com> Co-authored-by: Kemal <kemalmeler@outlook.com> Co-authored-by: Lazy_Frog <4590648+lazyFrogLOL@users.noreply.github.com> Co-authored-by: Yi Xiao <54782454+YIXIAO0@users.noreply.github.com> Co-authored-by: Steven sun <98230804+Tuyohai@users.noreply.github.com> Co-authored-by: steven <sunzwj@digitalchina.com> Co-authored-by: Kalo Chin <91766386+fdb02983rhy@users.noreply.github.com> Co-authored-by: Katy Tao <34019945+KatyTao@users.noreply.github.com> Co-authored-by: depy <42985524+h4ckdepy@users.noreply.github.com> Co-authored-by: 胡春东 <gycm520@gmail.com> Co-authored-by: Junjie.M <118170653@qq.com> Co-authored-by: MuYu <mr.muzea@gmail.com> Co-authored-by: Naoki Takashima <39912547+takatea@users.noreply.github.com> Co-authored-by: Summer-Gu <37869445+gubinjie@users.noreply.github.com> Co-authored-by: Fei He <droxer.he@gmail.com> Co-authored-by: ybalbert001 <120714773+ybalbert001@users.noreply.github.com> Co-authored-by: Yuanbo Li <ybalbert@amazon.com> Co-authored-by: douxc <7553076+douxc@users.noreply.github.com> Co-authored-by: liuzhenghua <1090179900@qq.com> Co-authored-by: Wu Jiayang <62842862+Wu-Jiayang@users.noreply.github.com> Co-authored-by: Your Name <you@example.com> Co-authored-by: kimjion <45935338+kimjion@users.noreply.github.com> Co-authored-by: AugNSo <song.tiankai@icloud.com> Co-authored-by: llinvokerl <38915183+llinvokerl@users.noreply.github.com> Co-authored-by: liusurong.lsr <liusurong.lsr@alibaba-inc.com> Co-authored-by: Vasu Negi <vasu-negi@users.noreply.github.com> Co-authored-by: Hundredwz <1808096180@qq.com> Co-authored-by: Xiyuan Chen <52963600+GareArc@users.noreply.github.com>
2025-12-25 02:33:00 +08:00 · 2025-02-17 17:05:13 +08:00
parent 222df44d21
commit 403e2d58b9
3272 changed files with 66339 additions and 281594 deletions
--- a/api/core/plugin/backwards_invocation/app.py
+++ b/api/core/plugin/backwards_invocation/app.py
@@ -0,0 +1,189 @@
+from collections.abc import Generator, Mapping
+from typing import Optional, Union
+
+from controllers.service_api.wraps import create_or_update_end_user_for_user_id
+from core.app.apps.advanced_chat.app_generator import AdvancedChatAppGenerator
+from core.app.apps.agent_chat.app_generator import AgentChatAppGenerator
+from core.app.apps.chat.app_generator import ChatAppGenerator
+from core.app.apps.completion.app_generator import CompletionAppGenerator
+from core.app.apps.workflow.app_generator import WorkflowAppGenerator
+from core.app.entities.app_invoke_entities import InvokeFrom
+from core.plugin.backwards_invocation.base import BaseBackwardsInvocation
+from extensions.ext_database import db
+from models.account import Account
+from models.model import App, AppMode, EndUser
+
+
+class PluginAppBackwardsInvocation(BaseBackwardsInvocation):
+    @classmethod
+    def invoke_app(
+        cls,
+        app_id: str,
+        user_id: str,
+        tenant_id: str,
+        conversation_id: Optional[str],
+        query: Optional[str],
+        stream: bool,
+        inputs: Mapping,
+        files: list[dict],
+    ) -> Generator[Mapping | str, None, None] | Mapping:
+        """
+        invoke app
+        """
+        app = cls._get_app(app_id, tenant_id)
+        if not user_id:
+            user = create_or_update_end_user_for_user_id(app)
+        else:
+            user = cls._get_user(user_id)
+
+        conversation_id = conversation_id or ""
+
+        if app.mode in {AppMode.ADVANCED_CHAT.value, AppMode.AGENT_CHAT.value, AppMode.CHAT.value}:
+            if not query:
+                raise ValueError("missing query")
+
+            return cls.invoke_chat_app(app, user, conversation_id, query, stream, inputs, files)
+        elif app.mode == AppMode.WORKFLOW.value:
+            return cls.invoke_workflow_app(app, user, stream, inputs, files)
+        elif app.mode == AppMode.COMPLETION:
+            return cls.invoke_completion_app(app, user, stream, inputs, files)
+
+        raise ValueError("unexpected app type")
+
+    @classmethod
+    def invoke_chat_app(
+        cls,
+        app: App,
+        user: Account | EndUser,
+        conversation_id: str,
+        query: str,
+        stream: bool,
+        inputs: Mapping,
+        files: list[dict],
+    ) -> Generator[Mapping | str, None, None] | Mapping:
+        """
+        invoke chat app
+        """
+        if app.mode == AppMode.ADVANCED_CHAT.value:
+            workflow = app.workflow
+            if not workflow:
+                raise ValueError("unexpected app type")
+
+            return AdvancedChatAppGenerator().generate(
+                app_model=app,
+                workflow=workflow,
+                user=user,
+                args={
+                    "inputs": inputs,
+                    "query": query,
+                    "files": files,
+                    "conversation_id": conversation_id,
+                },
+                invoke_from=InvokeFrom.SERVICE_API,
+                streaming=stream,
+            )
+        elif app.mode == AppMode.AGENT_CHAT.value:
+            return AgentChatAppGenerator().generate(
+                app_model=app,
+                user=user,
+                args={
+                    "inputs": inputs,
+                    "query": query,
+                    "files": files,
+                    "conversation_id": conversation_id,
+                },
+                invoke_from=InvokeFrom.SERVICE_API,
+                streaming=stream,
+            )
+        elif app.mode == AppMode.CHAT.value:
+            return ChatAppGenerator().generate(
+                app_model=app,
+                user=user,
+                args={
+                    "inputs": inputs,
+                    "query": query,
+                    "files": files,
+                    "conversation_id": conversation_id,
+                },
+                invoke_from=InvokeFrom.SERVICE_API,
+                streaming=stream,
+            )
+        else:
+            raise ValueError("unexpected app type")
+
+    @classmethod
+    def invoke_workflow_app(
+        cls,
+        app: App,
+        user: EndUser | Account,
+        stream: bool,
+        inputs: Mapping,
+        files: list[dict],
+    ) -> Generator[Mapping | str, None, None] | Mapping:
+        """
+        invoke workflow app
+        """
+        workflow = app.workflow
+        if not workflow:
+            raise ValueError("")
+
+        return WorkflowAppGenerator().generate(
+            app_model=app,
+            workflow=workflow,
+            user=user,
+            args={"inputs": inputs, "files": files},
+            invoke_from=InvokeFrom.SERVICE_API,
+            streaming=stream,
+            call_depth=1,
+            workflow_thread_pool_id=None,
+        )
+
+    @classmethod
+    def invoke_completion_app(
+        cls,
+        app: App,
+        user: EndUser | Account,
+        stream: bool,
+        inputs: Mapping,
+        files: list[dict],
+    ) -> Generator[Mapping | str, None, None] | Mapping:
+        """
+        invoke completion app
+        """
+        return CompletionAppGenerator().generate(
+            app_model=app,
+            user=user,
+            args={"inputs": inputs, "files": files},
+            invoke_from=InvokeFrom.SERVICE_API,
+            streaming=stream,
+        )
+
+    @classmethod
+    def _get_user(cls, user_id: str) -> Union[EndUser, Account]:
+        """
+        get the user by user id
+        """
+
+        user = db.session.query(EndUser).filter(EndUser.id == user_id).first()
+        if not user:
+            user = db.session.query(Account).filter(Account.id == user_id).first()
+
+        if not user:
+            raise ValueError("user not found")
+
+        return user
+
+    @classmethod
+    def _get_app(cls, app_id: str, tenant_id: str) -> App:
+        """
+        get app
+        """
+        try:
+            app = db.session.query(App).filter(App.id == app_id).filter(App.tenant_id == tenant_id).first()
+        except Exception:
+            raise ValueError("app not found")
+
+        if not app:
+            raise ValueError("app not found")
+
+        return app
--- a/api/core/plugin/backwards_invocation/base.py
+++ b/api/core/plugin/backwards_invocation/base.py
@@ -0,0 +1,29 @@
+from collections.abc import Generator, Mapping
+from typing import Generic, Optional, TypeVar
+
+from pydantic import BaseModel
+
+
+class BaseBackwardsInvocation:
+    @classmethod
+    def convert_to_event_stream(cls, response: Generator[BaseModel | Mapping | str, None, None] | BaseModel | Mapping):
+        if isinstance(response, Generator):
+            try:
+                for chunk in response:
+                    if isinstance(chunk, BaseModel | dict):
+                        yield BaseBackwardsInvocationResponse(data=chunk).model_dump_json().encode() + b"\n\n"
+                    elif isinstance(chunk, str):
+                        yield f"event: {chunk}\n\n".encode()
+            except Exception as e:
+                error_message = BaseBackwardsInvocationResponse(error=str(e)).model_dump_json()
+                yield f"{error_message}\n\n".encode()
+        else:
+            yield BaseBackwardsInvocationResponse(data=response).model_dump_json().encode() + b"\n\n"
+
+
+T = TypeVar("T", bound=dict | Mapping | str | bool | int | BaseModel)
+
+
+class BaseBackwardsInvocationResponse(BaseModel, Generic[T]):
+    data: Optional[T] = None
+    error: str = ""
--- a/api/core/plugin/backwards_invocation/encrypt.py
+++ b/api/core/plugin/backwards_invocation/encrypt.py
@@ -0,0 +1,30 @@
+from core.plugin.entities.request import RequestInvokeEncrypt
+from core.tools.utils.configuration import ProviderConfigEncrypter
+from models.account import Tenant
+
+
+class PluginEncrypter:
+    @classmethod
+    def invoke_encrypt(cls, tenant: Tenant, payload: RequestInvokeEncrypt) -> dict:
+        encrypter = ProviderConfigEncrypter(
+            tenant_id=tenant.id,
+            config=payload.config,
+            provider_type=payload.namespace,
+            provider_identity=payload.identity,
+        )
+
+        if payload.opt == "encrypt":
+            return {
+                "data": encrypter.encrypt(payload.data),
+            }
+        elif payload.opt == "decrypt":
+            return {
+                "data": encrypter.decrypt(payload.data),
+            }
+        elif payload.opt == "clear":
+            encrypter.delete_tool_credentials_cache()
+            return {
+                "data": {},
+            }
+        else:
+            raise ValueError(f"Invalid opt: {payload.opt}")
--- a/api/core/plugin/backwards_invocation/model.py
+++ b/api/core/plugin/backwards_invocation/model.py
@@ -0,0 +1,321 @@
+import tempfile
+from binascii import hexlify, unhexlify
+from collections.abc import Generator
+
+from core.model_manager import ModelManager
+from core.model_runtime.entities.llm_entities import LLMResult, LLMResultChunk
+from core.model_runtime.entities.message_entities import (
+    PromptMessage,
+    SystemPromptMessage,
+    UserPromptMessage,
+)
+from core.plugin.backwards_invocation.base import BaseBackwardsInvocation
+from core.plugin.entities.request import (
+    RequestInvokeLLM,
+    RequestInvokeModeration,
+    RequestInvokeRerank,
+    RequestInvokeSpeech2Text,
+    RequestInvokeSummary,
+    RequestInvokeTextEmbedding,
+    RequestInvokeTTS,
+)
+from core.tools.entities.tool_entities import ToolProviderType
+from core.tools.utils.model_invocation_utils import ModelInvocationUtils
+from core.workflow.nodes.llm.node import LLMNode
+from models.account import Tenant
+
+
+class PluginModelBackwardsInvocation(BaseBackwardsInvocation):
+    @classmethod
+    def invoke_llm(
+        cls, user_id: str, tenant: Tenant, payload: RequestInvokeLLM
+    ) -> Generator[LLMResultChunk, None, None] | LLMResult:
+        """
+        invoke llm
+        """
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant.id,
+            provider=payload.provider,
+            model_type=payload.model_type,
+            model=payload.model,
+        )
+
+        # invoke model
+        response = model_instance.invoke_llm(
+            prompt_messages=payload.prompt_messages,
+            model_parameters=payload.completion_params,
+            tools=payload.tools,
+            stop=payload.stop,
+            stream=payload.stream or True,
+            user=user_id,
+        )
+
+        if isinstance(response, Generator):
+
+            def handle() -> Generator[LLMResultChunk, None, None]:
+                for chunk in response:
+                    if chunk.delta.usage:
+                        LLMNode.deduct_llm_quota(
+                            tenant_id=tenant.id, model_instance=model_instance, usage=chunk.delta.usage
+                        )
+                    yield chunk
+
+            return handle()
+        else:
+            if response.usage:
+                LLMNode.deduct_llm_quota(tenant_id=tenant.id, model_instance=model_instance, usage=response.usage)
+            return response
+
+    @classmethod
+    def invoke_text_embedding(cls, user_id: str, tenant: Tenant, payload: RequestInvokeTextEmbedding):
+        """
+        invoke text embedding
+        """
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant.id,
+            provider=payload.provider,
+            model_type=payload.model_type,
+            model=payload.model,
+        )
+
+        # invoke model
+        response = model_instance.invoke_text_embedding(
+            texts=payload.texts,
+            user=user_id,
+        )
+
+        return response
+
+    @classmethod
+    def invoke_rerank(cls, user_id: str, tenant: Tenant, payload: RequestInvokeRerank):
+        """
+        invoke rerank
+        """
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant.id,
+            provider=payload.provider,
+            model_type=payload.model_type,
+            model=payload.model,
+        )
+
+        # invoke model
+        response = model_instance.invoke_rerank(
+            query=payload.query,
+            docs=payload.docs,
+            score_threshold=payload.score_threshold,
+            top_n=payload.top_n,
+            user=user_id,
+        )
+
+        return response
+
+    @classmethod
+    def invoke_tts(cls, user_id: str, tenant: Tenant, payload: RequestInvokeTTS):
+        """
+        invoke tts
+        """
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant.id,
+            provider=payload.provider,
+            model_type=payload.model_type,
+            model=payload.model,
+        )
+
+        # invoke model
+        response = model_instance.invoke_tts(
+            content_text=payload.content_text,
+            tenant_id=tenant.id,
+            voice=payload.voice,
+            user=user_id,
+        )
+
+        def handle() -> Generator[dict, None, None]:
+            for chunk in response:
+                yield {"result": hexlify(chunk).decode("utf-8")}
+
+        return handle()
+
+    @classmethod
+    def invoke_speech2text(cls, user_id: str, tenant: Tenant, payload: RequestInvokeSpeech2Text):
+        """
+        invoke speech2text
+        """
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant.id,
+            provider=payload.provider,
+            model_type=payload.model_type,
+            model=payload.model,
+        )
+
+        # invoke model
+        with tempfile.NamedTemporaryFile(suffix=".mp3", mode="wb", delete=True) as temp:
+            temp.write(unhexlify(payload.file))
+            temp.flush()
+            temp.seek(0)
+
+            response = model_instance.invoke_speech2text(
+                file=temp,
+                user=user_id,
+            )
+
+            return {
+                "result": response,
+            }
+
+    @classmethod
+    def invoke_moderation(cls, user_id: str, tenant: Tenant, payload: RequestInvokeModeration):
+        """
+        invoke moderation
+        """
+        model_instance = ModelManager().get_model_instance(
+            tenant_id=tenant.id,
+            provider=payload.provider,
+            model_type=payload.model_type,
+            model=payload.model,
+        )
+
+        # invoke model
+        response = model_instance.invoke_moderation(
+            text=payload.text,
+            user=user_id,
+        )
+
+        return {
+            "result": response,
+        }
+
+    @classmethod
+    def get_system_model_max_tokens(cls, tenant_id: str) -> int:
+        """
+        get system model max tokens
+        """
+        return ModelInvocationUtils.get_max_llm_context_tokens(tenant_id=tenant_id)
+
+    @classmethod
+    def get_prompt_tokens(cls, tenant_id: str, prompt_messages: list[PromptMessage]) -> int:
+        """
+        get prompt tokens
+        """
+        return ModelInvocationUtils.calculate_tokens(tenant_id=tenant_id, prompt_messages=prompt_messages)
+
+    @classmethod
+    def invoke_system_model(
+        cls,
+        user_id: str,
+        tenant: Tenant,
+        prompt_messages: list[PromptMessage],
+    ) -> LLMResult:
+        """
+        invoke system model
+        """
+        return ModelInvocationUtils.invoke(
+            user_id=user_id,
+            tenant_id=tenant.id,
+            tool_type=ToolProviderType.PLUGIN,
+            tool_name="plugin",
+            prompt_messages=prompt_messages,
+        )
+
+    @classmethod
+    def invoke_summary(cls, user_id: str, tenant: Tenant, payload: RequestInvokeSummary):
+        """
+        invoke summary
+        """
+        max_tokens = cls.get_system_model_max_tokens(tenant_id=tenant.id)
+        content = payload.text
+
+        SUMMARY_PROMPT = """You are a professional language researcher, you are interested in the language
+and you can quickly aimed at the main point of an webpage and reproduce it in your own words but 
+retain the original meaning and keep the key points. 
+however, the text you got is too long, what you got is possible a part of the text.
+Please summarize the text you got.
+
+Here is the extra instruction you need to follow:
+<extra_instruction>
+{payload.instruction}
+</extra_instruction>
+"""
+
+        if (
+            cls.get_prompt_tokens(
+                tenant_id=tenant.id,
+                prompt_messages=[UserPromptMessage(content=content)],
+            )
+            < max_tokens * 0.6
+        ):
+            return content
+
+        def get_prompt_tokens(content: str) -> int:
+            return cls.get_prompt_tokens(
+                tenant_id=tenant.id,
+                prompt_messages=[
+                    SystemPromptMessage(content=SUMMARY_PROMPT.replace("{payload.instruction}", payload.instruction)),
+                    UserPromptMessage(content=content),
+                ],
+            )
+
+        def summarize(content: str) -> str:
+            summary = cls.invoke_system_model(
+                user_id=user_id,
+                tenant=tenant,
+                prompt_messages=[
+                    SystemPromptMessage(content=SUMMARY_PROMPT.replace("{payload.instruction}", payload.instruction)),
+                    UserPromptMessage(content=content),
+                ],
+            )
+
+            assert isinstance(summary.message.content, str)
+            return summary.message.content
+
+        lines = content.split("\n")
+        new_lines: list[str] = []
+        # split long line into multiple lines
+        for i in range(len(lines)):
+            line = lines[i]
+            if not line.strip():
+                continue
+            if len(line) < max_tokens * 0.5:
+                new_lines.append(line)
+            elif get_prompt_tokens(line) > max_tokens * 0.7:
+                while get_prompt_tokens(line) > max_tokens * 0.7:
+                    new_lines.append(line[: int(max_tokens * 0.5)])
+                    line = line[int(max_tokens * 0.5) :]
+                new_lines.append(line)
+            else:
+                new_lines.append(line)
+
+        # merge lines into messages with max tokens
+        messages: list[str] = []
+        for i in new_lines:  # type: ignore
+            if len(messages) == 0:
+                messages.append(i)  # type: ignore
+            else:
+                if len(messages[-1]) + len(i) < max_tokens * 0.5:  # type: ignore
+                    messages[-1] += i  # type: ignore
+                if get_prompt_tokens(messages[-1] + i) > max_tokens * 0.7:  # type: ignore
+                    messages.append(i)  # type: ignore
+                else:
+                    messages[-1] += i  # type: ignore
+
+        summaries = []
+        for i in range(len(messages)):
+            message = messages[i]
+            summary = summarize(message)
+            summaries.append(summary)
+
+        result = "\n".join(summaries)
+
+        if (
+            cls.get_prompt_tokens(
+                tenant_id=tenant.id,
+                prompt_messages=[UserPromptMessage(content=result)],
+            )
+            > max_tokens * 0.7
+        ):
+            return cls.invoke_summary(
+                user_id=user_id,
+                tenant=tenant,
+                payload=RequestInvokeSummary(text=result, instruction=payload.instruction),
+            )
+
+        return result
--- a/api/core/plugin/backwards_invocation/node.py
+++ b/api/core/plugin/backwards_invocation/node.py
@@ -0,0 +1,117 @@
+from core.plugin.backwards_invocation.base import BaseBackwardsInvocation
+from core.workflow.nodes.enums import NodeType
+from core.workflow.nodes.parameter_extractor.entities import (
+    ModelConfig as ParameterExtractorModelConfig,
+)
+from core.workflow.nodes.parameter_extractor.entities import (
+    ParameterConfig,
+    ParameterExtractorNodeData,
+)
+from core.workflow.nodes.question_classifier.entities import (
+    ClassConfig,
+    QuestionClassifierNodeData,
+)
+from core.workflow.nodes.question_classifier.entities import (
+    ModelConfig as QuestionClassifierModelConfig,
+)
+from services.workflow_service import WorkflowService
+
+
+class PluginNodeBackwardsInvocation(BaseBackwardsInvocation):
+    @classmethod
+    def invoke_parameter_extractor(
+        cls,
+        tenant_id: str,
+        user_id: str,
+        parameters: list[ParameterConfig],
+        model_config: ParameterExtractorModelConfig,
+        instruction: str,
+        query: str,
+    ) -> dict:
+        """
+        Invoke parameter extractor node.
+
+        :param tenant_id: str
+        :param user_id: str
+        :param parameters: list[ParameterConfig]
+        :param model_config: ModelConfig
+        :param instruction: str
+        :param query: str
+        :return: dict
+        """
+        workflow_service = WorkflowService()
+        node_id = "1919810"
+        node_data = ParameterExtractorNodeData(
+            title="parameter_extractor",
+            desc="parameter_extractor",
+            parameters=parameters,
+            reasoning_mode="function_call",
+            query=[node_id, "query"],
+            model=model_config,
+            instruction=instruction,  # instruct with variables are not supported
+        )
+        node_data_dict = node_data.model_dump()
+        node_data_dict["type"] = NodeType.PARAMETER_EXTRACTOR.value
+        execution = workflow_service.run_free_workflow_node(
+            node_data_dict,
+            tenant_id=tenant_id,
+            user_id=user_id,
+            node_id=node_id,
+            user_inputs={
+                f"{node_id}.query": query,
+            },
+        )
+
+        return {
+            "inputs": execution.inputs_dict,
+            "outputs": execution.outputs_dict,
+            "process_data": execution.process_data_dict,
+        }
+
+    @classmethod
+    def invoke_question_classifier(
+        cls,
+        tenant_id: str,
+        user_id: str,
+        model_config: QuestionClassifierModelConfig,
+        classes: list[ClassConfig],
+        instruction: str,
+        query: str,
+    ) -> dict:
+        """
+        Invoke question classifier node.
+
+        :param tenant_id: str
+        :param user_id: str
+        :param model_config: ModelConfig
+        :param classes: list[ClassConfig]
+        :param instruction: str
+        :param query: str
+        :return: dict
+        """
+        workflow_service = WorkflowService()
+        node_id = "1919810"
+        node_data = QuestionClassifierNodeData(
+            title="question_classifier",
+            desc="question_classifier",
+            query_variable_selector=[node_id, "query"],
+            model=model_config,
+            classes=classes,
+            instruction=instruction,  # instruct with variables are not supported
+        )
+        node_data_dict = node_data.model_dump()
+        execution = workflow_service.run_free_workflow_node(
+            node_data_dict,
+            tenant_id=tenant_id,
+            user_id=user_id,
+            node_id=node_id,
+            user_inputs={
+                f"{node_id}.query": query,
+            },
+        )
+
+        return {
+            "inputs": execution.inputs_dict,
+            "outputs": execution.outputs_dict,
+            "process_data": execution.process_data_dict,
+        }
--- a/api/core/plugin/backwards_invocation/tool.py
+++ b/api/core/plugin/backwards_invocation/tool.py
@@ -0,0 +1,45 @@
+from collections.abc import Generator
+from typing import Any
+
+from core.callback_handler.workflow_tool_callback_handler import DifyWorkflowCallbackHandler
+from core.plugin.backwards_invocation.base import BaseBackwardsInvocation
+from core.tools.entities.tool_entities import ToolInvokeMessage, ToolProviderType
+from core.tools.tool_engine import ToolEngine
+from core.tools.tool_manager import ToolManager
+from core.tools.utils.message_transformer import ToolFileMessageTransformer
+
+
+class PluginToolBackwardsInvocation(BaseBackwardsInvocation):
+    """
+    Backwards invocation for plugin tools.
+    """
+
+    @classmethod
+    def invoke_tool(
+        cls,
+        tenant_id: str,
+        user_id: str,
+        tool_type: ToolProviderType,
+        provider: str,
+        tool_name: str,
+        tool_parameters: dict[str, Any],
+    ) -> Generator[ToolInvokeMessage, None, None]:
+        """
+        invoke tool
+        """
+        # get tool runtime
+        try:
+            tool_runtime = ToolManager.get_tool_runtime_from_plugin(
+                tool_type, tenant_id, provider, tool_name, tool_parameters
+            )
+            response = ToolEngine.generic_invoke(
+                tool_runtime, tool_parameters, user_id, DifyWorkflowCallbackHandler(), workflow_call_depth=1
+            )
+
+            response = ToolFileMessageTransformer.transform_tool_invoke_messages(
+                response, user_id=user_id, tenant_id=tenant_id
+            )
+
+            return response
+        except Exception as e:
+            raise e