Skip to content

Built-in Node Catalog

Orcheo currently ships with 138 built-in nodes across 21 categories.

This catalog is sourced from runtime node registry metadata. Run orcheo node list to inspect the nodes available in your environment, including custom registrations.

Summary

Category Node Count
AI (ai) 4
Agentensor (agentensor) 1
Base (base) 1
Browser (browser) 6
Communication (communication) 2
Conversational Search (conversational_search) 46
Data (data) 6
Evaluation (evaluation) 8
Lark (lark) 2
LinkedIn (linkedin) 1
Logic (logic) 3
Messaging (messaging) 4
MongoDB (mongodb) 9
RSS (rss) 1
Slack (slack) 2
Storage (storage) 3
Telegram (telegram) 1
Trigger (trigger) 7
Utility (utility) 5
WeCom (wecom) 9
Workflow (workflow) 17

AI Nodes

Node Description
AgentNode Execute an AI agent with tools
AgentReplyExtractorNode Extract the final assistant reply from agent messages
DeepAgentNode Execute an autonomous deep-research agent with configurable iteration depth for multi-step tool use and synthesis
LLMNode Execute a text-only LLM call

Agentensor Nodes

Node Description
AgentensorNode Evaluate or train agent prompts using Agentensor datasets and evaluators.

Base Nodes

Node Description
NoOpTaskNode A no-op node for developers to use as a template for custom nodes. Do not use this node directly, but inherit from this with your own run method.

Browser Nodes

Node Description
BrowserActionNode Perform a user-style action against the active Playwright page.
BrowserCloseNode Close an active Playwright browser session.
BrowserExtractNode Extract page or element content from the active Playwright page.
BrowserNavigateNode Open a URL in a shared Playwright browser session.
BrowserScriptNode Evaluate JavaScript in the active Playwright page context.
BrowserWaitNode Wait for a condition on the active Playwright page.

Communication Nodes

Node Description
DiscordWebhookNode Send messages to Discord via incoming webhooks.
EmailNode Send an email via SMTP with optional TLS and authentication.

Conversational Search Nodes

Node Description
ABTestingNode Rank variants and gate rollouts using evaluation metrics.
AnalyticsExportNode Aggregate evaluation metrics and feedback for export.
AnswerCachingNode Cache answers by query with TTL-based eviction.
AnswerQualityEvaluationNode Score generated answers against reference answers.
ChunkEmbeddingNode Generate vector records for document chunks via configurable embedding functions.
ChunkingStrategyNode Split documents into overlapping chunks for indexing.
CitationsFormatterNode Format citation metadata into human-readable strings.
ContextCompressorNode Summarize retrieved context using an AI model so downstream nodes can consume a condensed evidence block.
ConversationCompressorNode Summarize and budget a conversation history for downstream use.
ConversationStateNode Load and persist conversation history for a session.
CoreferenceResolverNode Resolve simple pronouns using prior conversation turns.
DataAugmentationNode Generate synthetic variants of dataset entries.
DatasetNode Load and filter golden datasets for evaluation workflows.
DenseSearchNode Perform embedding-based retrieval via a configured vector store.
DocumentLoaderNode Normalize raw document payloads into validated Document objects.
FailureAnalysisNode Categorize evaluation failures for triage.
FeedbackIngestionNode Persist feedback entries with deduplication.
GroundedGeneratorNode Generate grounded answers with citations and retry semantics.
HallucinationGuardNode Validate generator output for citations and completeness.
HybridFusionNode Fuse results from multiple retrievers using RRF or weighted sum.
IncrementalIndexerNode Index or update chunks incrementally with retry and backpressure controls.
LLMJudgeNode Apply lightweight, AI model judging heuristics.
MemoryPrivacyNode Enforce redaction and retention for conversation history.
MemorySummarizerNode Persist a compact conversation summary into the memory store.
MetadataExtractorNode Attach structured metadata to normalized documents.
MultiHopPlannerNode Derive sequential sub-queries for multi-hop answering.
PineconeRerankNode Rerank retrieval results via Pinecone inference for tighter ordering.
PolicyComplianceNode Apply policy checks and emit audit details.
QueryClarificationNode Generate clarifying prompts when ambiguity is detected.
QueryClassifierNode Classify a query intent to support routing decisions.
QueryRewriteNode Rewrite or expand a query using recent conversation context to improve recall.
ReRankerNode Apply secondary scoring to retrieval results for better ranking.
RetrievalEvaluationNode Compute retrieval quality metrics for search results.
SearchResultAdapterNode Normalize arbitrary retrieval payloads into SearchResult items.
SearchResultFormatterNode Format SearchResult entries into markdown for tool responses.
SessionManagementNode Manage conversation sessions with capacity controls.
SourceRouterNode Route fused results into per-source buckets with filtering.
SparseSearchNode Perform sparse keyword retrieval using BM25 scoring.
StreamingGeneratorNode Generate responses and stream token chunks with backpressure.
TextEmbeddingNode Embed one or more text inputs using a configurable embedding model.
TopicShiftDetectorNode Detect whether a new query diverges from recent conversation context.
TurnAnnotationNode Annotate conversation turns with heuristics.
UserFeedbackCollectionNode Normalize and validate explicit user feedback.
VectorStoreUpsertNode Persist vector records produced by an embedding node into storage.
WebDocumentLoaderNode Fetch web pages and convert them to normalized Document objects.
WebSearchNode Perform live web search via the Tavily API.

Data Nodes

Node Description
CodeNode Base class for user-authored node logic; the sole customisation port. Do not use this node directly, but inherit from this with your own run method.
DataTransformNode Map values from an input payload into a transformed structure.
HtmlTextTransformNode Decode, escape, and normalise HTML text in structured data.
HttpRequestNode Perform an HTTP request and return the response payload.
JsonProcessingNode Parse, stringify, or extract data from JSON payloads.
MergeNode Merge multiple payloads into a single aggregate structure.

Evaluation Nodes

Node Description
BleuMetricsNode Compute SacreBLEU between predicted and reference texts
ConversationalBatchEvalNode Iterate conversations and turns through a pipeline, collecting predictions paired with gold labels
MultiDoc2DialCorpusLoaderNode Load MultiDoc2Dial corpus documents from a local path or URL and normalize them for indexing.
MultiDoc2DialDatasetNode Load MultiDoc2Dial conversations with gold responses for evaluation
QReCCDatasetNode Load QReCC conversations with gold rewrites for evaluation
RougeMetricsNode Compute ROUGE scores between predicted and reference texts
SemanticSimilarityMetricsNode Compute embedding cosine similarity between predicted and reference texts
TokenF1MetricsNode Compute token-level F1 between predicted and reference texts

Lark Nodes

Node Description
LarkSendMessageNode Send messages to Lark chats or reply to inbound Lark messages
LarkTenantAccessTokenNode Fetch a tenant access token from the Lark auth API

LinkedIn Nodes

Node Description
LinkedInPostNode Create a LinkedIn post using vault-backed access credentials.

Logic Nodes

Node Description
ExtractAIMessageNode Extract a text AI message from a structured response
HumanInputNode Pause a workflow and collect human input
StructuredRouterDispatchNode Translate structured agent output into a routing result

Messaging Nodes

Node Description
MessageDiscordNode Send a message to a Discord channel using a bot token.
MessageQQNode Send a message to QQ using AppID and client secret.
MessageTelegramNode Send message to Telegram
TelegramSendDocumentNode Send a document file to Telegram

MongoDB Nodes

Node Description
MongoDBAggregateNode MongoDB aggregate wrapper
MongoDBEnsureSearchIndexNode Ensure a MongoDB Atlas Search index exists.
MongoDBEnsureVectorIndexNode Ensure a MongoDB Atlas vector search index exists.
MongoDBFindNode MongoDB find wrapper with sort and limit support
MongoDBHybridSearchNode Execute a hybrid search over text and vector indexes asynchronously.
MongoDBInsertManyNode Insert documents into MongoDB with optional vectors
MongoDBNode MongoDB node
MongoDBUpdateManyNode MongoDB update_many wrapper
MongoDBUpsertManyNode Bulk-upsert upstream records into MongoDB using keyed filters

RSS Nodes

Node Description
RSSNode Fetch and parse RSS/Atom feeds from multiple sources

Slack Nodes

Node Description
SlackEventsParserNode Validate and parse Slack Events API payloads
SlackNode Slack node

Storage Nodes

Node Description
FileUploadNode Upload a single file to blob storage and return a download URL. Reads content, filename, and mime_type from state via template references and writes attachment_id and download_url to node_results.
GraphStoreAppendMessageNode Append a message to a versioned chat history item in the LangGraph graph store.
PostgresNode Execute SQL against a PostgreSQL database using psycopg.

Telegram Nodes

Node Description
TelegramEventsParserNode Validate and parse Telegram Bot webhook updates

Trigger Nodes

Node Description
CronTriggerNode Configure a cron-based schedule trigger.
DiscordBotListenerNode Receive Discord bot messages through the Gateway.
HttpPollingTriggerNode Poll an HTTP endpoint on an interval to trigger runs.
ManualTriggerNode Trigger workflows manually from the dashboard.
QQBotListenerNode Receive QQ bot messages through the managed Gateway.
TelegramBotListenerNode Receive Telegram bot updates through managed long polling.
WebhookTriggerNode Configure an HTTP webhook trigger.

Utility Nodes

Node Description
DebugNode Capture state snapshots and emit debug information.
DelayNode Pause execution for a fixed duration
ForLoopNode Iterate over a list, exposing one item per invocation
SetVariableNode Store variables for downstream nodes
SubWorkflowNode Execute a mini workflow inline using the node registry.

WeCom Nodes

Node Description
WeComAIBotEventsParserNode Validate WeCom AI bot signatures and parse callbacks
WeComAIBotPassiveReplyNode Encrypt and return passive AI bot replies
WeComAIBotResponseNode Send active replies to WeCom AI bot response_url
WeComAccessTokenNode Fetch and cache WeCom access token
WeComCustomerServiceSendNode Send messages via WeCom Customer Service (微信客服)
WeComCustomerServiceSyncNode Sync messages from WeCom Customer Service (微信客服)
WeComEventsParserNode Validate WeCom signatures and parse callback payloads
WeComGroupPushNode Send messages to WeCom group via webhook
WeComSendMessageNode Send messages to WeCom chat

Workflow Nodes

Node Description
CodebookOutputNode Render the produced codebook as Markdown
CodedDataIngestNode Parse a coded-data CSV into units, assignments, quantification
DataQualityNode Flag low-effort, duplicate, PII, and AI-like responses
ExportCodebookNode Export the current codebook to CSV or inline JSON text
ExportCodedDataNode Serialise coded units and assignments to a CSV download
ExportReportNode Regenerate the downloadable Markdown report link
IngestNode Parse the source payload into units
InsightCriticNode Find counter-evidence and annotate candidate insights
LLMStageFinalizeNode Persist a qualitative LLM stage response
LLMStagePrepareNode Prepare prompt payloads for a qualitative LLM stage
LoadAttachmentsNode Resolve uploaded attachments into readable payloads
RecodeOutputNode Render recoded data with a coded-data CSV download link
RecommendationGeneratorNode Attach Finding -> Action -> Expected impact recommendations
ReportOutputNode Render the final report and return it with a download link
SegmentRecordsNode Segment structured records into qualitative coding units
TwoTrackThemeReportNode Render a Covered/Emergent salience-ranked theme report
ValidateFilesNode Validate qualitative data files and optional codebooks

Creating Custom Nodes

See the Plugin Author Guide for instructions on creating and registering plugin-managed nodes.