langchain-core-workflow-a
Build LangChain chains and prompts for structured LLM workflows. Use when creating prompt templates, building LCEL chains, or implementing sequential processing pipelines. Trigger with phrases like "langchain chains", "langchain prompts", "LCEL workflow", "langchain pipeline", "prompt template".
- License 许可
- MIT
- 最近更新
- 2026-11-15
- Stars
- ★ 2.0k
- Forks
- 233
01 前言 Front Matter
- name
- langchain-core-workflow-a
- author
- Jeremy Longshore <jeremy@intentsolutions.io>
- license
- MIT
- version
- 1.0.0
- description
- Build LangChain chains and prompts for structured LLM workflows. Use when creating prompt templates, building LCEL chains, or implementing sequential processing pipelines. Trigger with phrases like "langchain chains", "langchain prompts", "LCEL workflow", "langchain pipeline", "prompt template".
- allowed-tools
- Read, Write, Edit
03 SKILL.md 正文
LangChain Core Workflow A: Chains & Prompts
Overview
Build production-ready chains using LangChain Expression Language (LCEL) with prompt templates, output parsers, and composition patterns.
Prerequisites
- Completed
langchain-install-authsetup - Understanding of prompt engineering basics
- Familiarity with Python type hints
Instructions
Step 1: Create Prompt Templates
from langchain_core.prompts import (
ChatPromptTemplate,
SystemMessagePromptTemplate,
HumanMessagePromptTemplate,
MessagesPlaceholder
)
# Simple template
simple_prompt = ChatPromptTemplate.from_template(
"Translate '{text}' to {language}"
)
# Chat-style template
chat_prompt = ChatPromptTemplate.from_messages([
SystemMessagePromptTemplate.from_template(
"You are a {role}. Respond in {style} style."
),
MessagesPlaceholder(variable_name="history", optional=True),
HumanMessagePromptTemplate.from_template("{input}")
])
Step 2: Build LCEL Chains
from langchain_openai import ChatOpenAI
from langchain_core.output_parsers import StrOutputParser, JsonOutputParser
llm = ChatOpenAI(model="gpt-4o-mini")
# Basic chain: prompt -> llm -> parser
basic_chain = simple_prompt | llm | StrOutputParser()
# Invoke the chain
result = basic_chain.invoke({
"text": "Hello, world!",
"language": "Spanish"
})
print(result) # "Hola, mundo!"
Step 3: Chain Composition
from langchain_core.runnables import RunnablePassthrough, RunnableParallel
# Sequential chain
chain1 = prompt1 | llm | StrOutputParser()
chain2 = prompt2 | llm | StrOutputParser()
sequential = chain1 | (lambda x: {"summary": x}) | chain2
# Parallel execution
parallel = RunnableParallel(
summary=prompt1 | llm | StrOutputParser(),
keywords=prompt2 | llm | StrOutputParser(),
sentiment=prompt3 | llm | StrOutputParser()
)
results = parallel.invoke({"text": "Your input text"})
# Returns: {"summary": "...", "keywords": "...", "sentiment": "..."}
Step 4: Branching Logic
from langchain_core.runnables import RunnableBranch
# Conditional branching
branch = RunnableBranch(
(lambda x: x["type"] == "question", question_chain),
(lambda x: x["type"] == "command", command_chain),
default_chain # Fallback
)
result = branch.invoke({"type": "question", "input": "What is AI?"})
Output
- Reusable prompt templates with variable substitution
- Type-safe LCEL chains with clear data flow
- Composable chain patterns (sequential, parallel, branching)
- Consistent output parsing
Examples
Multi-Step Processing Chain
from langchain_openai import ChatOpenAI
from langchain_core.prompts import ChatPromptTemplate
from langchain_core.output_parsers import StrOutputParser
llm = ChatOpenAI(model="gpt-4o-mini")
# Step 1: Extract key points
extract_prompt = ChatPromptTemplate.from_template(
"Extract 3 key points from: {text}"
)
# Step 2: Summarize
summarize_prompt = ChatPromptTemplate.from_template(
"Create a one-sentence summary from these points: {points}"
)
# Compose the chain
chain = (
{"points": extract_prompt | llm | StrOutputParser()}
| summarize_prompt
| llm
| StrOutputParser()
)
summary = chain.invoke({"text": "Long article text here..."})
With Context Injection
from langchain_core.runnables import RunnablePassthrough
def get_context(input_dict):
"""Fetch relevant context from database."""
return f"Context for: {input_dict['query']}"
chain = (
RunnablePassthrough.assign(context=get_context)
| prompt
| llm
| StrOutputParser()
)
result = chain.invoke({"query": "user question"})
Error Handling
| Error | Cause | Solution |
|---|---|---|
| Missing Variable | Template variable not provided | Check input dict keys match template |
| Type Error | Wrong input type | Ensure inputs match expected schema |
| Parse Error | Output doesn't match parser | Use more specific prompts or fallback |
Resources
Next Steps
Proceed to langchain-core-workflow-b for agents and tools workflow.
04 同仓库更多 Skills
theme-factory
Toolkit for styling artifacts with a theme. These artifacts can be slides, docs, reportings, HTML landing pages, etc. There are 10 pre-set themes with colors/fonts that you can apply to any artifact that has been creating, or can generate a new theme on-the-fly.
web-artifacts-builder
Suite of tools for creating elaborate, multi-component claude.ai HTML artifacts using modern frontend web technologies (React, Tailwind CSS, shadcn/ui). Use for complex artifacts requiring state management, routing, or shadcn/ui components - not for simple single-file HTML/JSX artifacts.
webapp-testing
Toolkit for interacting with and testing local web applications using Playwright. Supports verifying frontend functionality, debugging UI behavior, capturing browser screenshots, and viewing browser logs.
neuropixels-analysis
Neuropixels neural recording analysis. Load SpikeGLX/OpenEphys data, preprocess, motion correction, Kilosort4 spike sorting, quality metrics, Allen/IBL curation, AI-assisted visual analysis, for Neuropixels 1.0/2.0 extracellular electrophysiology. Use when working with neural recordings, spike sorting, extracellular electrophysiology, or when the user mentions Neuropixels, SpikeGLX, Open Ephys, Kilosort, quality metrics, or unit curation.
skill-creator
Guide for creating effective skills. This skill should be used when users want to create a new skill (or update an existing skill) that extends Claude's capabilities with specialized knowledge, workflows, or tool integrations.
notebooklm
Use this skill to query your Google NotebookLM notebooks directly from Claude Code for source-grounded, citation-backed answers from Gemini. Browser automation, library management, persistent auth. Drastically reduced hallucinations through document-only responses.