skills.rest
winsorllc
winsorllcCOMMUNITY
Explore all skills by this author →
GitHub

langgraph-agent

Creates LangGraph-based agents with automatic tool loops for OpenAI-compatible LLM providers.

1|Updated Feb 25, 2026
Search Tags:#automation#llm#agent#tool calling#task execution#langgraph
One-click install
npx skills add https://github.com/winsorllc/upgraded-carnival --skill langgraph-agent
Or copy as Structured Prompt for Agent▼
Please help me install this Agent Skill.
Skill: langgraph-agent
Source: https://github.com/winsorllc/upgraded-carnival/tree/main/.pi/skills/langgraph-agent
Command: npx skills add https://github.com/winsorllc/upgraded-carnival --skill langgraph-agent

SYSTEM DOCUMENTATION & REQUIREMENTS

💡 This Skill includes scripts (resource) components.

What problem does it solve?

This Skill provides a robust framework for executing multi-step tasks reliably, especially when dealing with LLMs that have inconsistent tool-calling capabilities.

Core Features & Use Cases

  • Consistent Tool Calling: Ensures tools are called correctly across various OpenAI-compatible LLM providers.
  • Automatic Tool Loops: Automatically retries tool calls until a task is successfully completed.
  • Unified Interface: Works with multiple LLM providers (Z.AI, OpenRouter, Groq, DeepSeek, Ollama) through a single interface.
  • Use Case: Building a custom agent that needs to research a topic, summarize findings, and then draft an email, ensuring each step is executed correctly even if the LLM struggles with tool invocation.

Quick Start

Create an agent with the 'shell' and 'file_read' tools and use it to execute the command 'ls -l'.

Frequently Asked Questions about langgraph-agent

High-intent search queries and answers about installing and using this skill.

FAQPage Schema
How do I execute multi-step tasks reliably with LLMs that have inconsistent tool calling?▼

Multi-step task execution requires a LangGraph-based agent framework that ensures consistent tool calling across OpenAI-compatible LLM providers. It uses automatic tool loops to retry tool calls until tasks finish successfully.

Can I use LangGraph to build an agent that works with Groq, DeepSeek, and Ollama?▼

Yes, you can use LangGraph to build a unified agent interface that works with Groq, DeepSeek, Ollama, Z.AI/GLM-5, and OpenRouter. You need an LLM API key and optionally a custom base URL and model name for integration.

What is the best way to automate a workflow that researches, summarizes, and drafts an email?▼

The best way to automate this multi-step workflow is using a LangGraph agent with automatic tool loops. This ensures each step like researching, summarizing, and drafting executes correctly even if the LLM struggles with tool invocation.

How do I set up a LangGraph agent to execute a shell command like 'ls -l'?▼

To execute a shell command, create a LangGraph agent equipped with the 'shell' and 'file_read' tools. The framework will handle the tool calling loop automatically to run the command like 'ls -l' and return the output.

Why does my LLM agent fail to call tools correctly across different providers?▼

LLM agents fail tool calls due to inconsistent tool-calling capabilities across different providers. A LangGraph-based framework solves this by standardizing tool calling and automatically retrying failed tool invocations.

Related Skills

Explore More Software Engineering Skills→
xiaomochnxiaomochn
198

model-config

Configure AI model providers and default model selection in openclaw.json.

Community
Intermediate
Xza85hrfXza85hrf

programmatic-tool-calling

Orchestrate Claude agents to invoke external tools in multi-step workflows.

Community
Advanced
FatByFatBy
2

file-operations

Read, write, append, and list local files via shell access.

Community
Basic
DemonDamonDemonDamon
207

agenticx-agent-builder

Design, configure, and deploy AgenticX agents with roles, goals, tools, and LLM providers for scalable multi-agent systems.

Community
Advanced