temporal-sa

Temporal Langchain Incident Demo — AI skill for Claude Code

AI community

An AI incident-response agent showing how Temporal and LangChain fit together: Temporal runs the durable outer loop, LangChain each agent turn.

How to install Temporal Langchain Incident Demo

This entry records only its repository, not the path inside it, so there is no exact command to give. Open temporal-sa/temporal-langchain-incident-demo and copy the folder into ~/.claude/skills/, or the file into ~/.claude/agents/.

What Temporal Langchain Incident Demo does

An AI incident-response agent showing how Temporal and LangChain fit together: Temporal runs the durable outer loop, LangChain each agent turn. Includes MCP tools, tools as activities, human-in-the-loop approval, a Java worker, and a live console built from event history. Works with Claude or any LangChain LLM.

Alternatives in AI

Anthropic Quickstarts — by Anthropic - Offers comprehensive development guides for three distinct AI-powered demo projects with standa 15.4k ★
Project Next 14 AI Prompt Sharing — Next.js recently became the official React framework as outlined in React docs 3k ★
Memtrace Public — Structural memory for AI coding agents 466 ★

README

Durable incident copilot

**An AI agent built with LangChain and MCP, running on Temporal, using Claude or any LangChain-supported LLM.** Start an incident, watch the agent investigate and fix it, then break things on purpose (kill the worker, take down a tool server, ship a bug) and watch it recover without losing its place.

**Temporal runs the outer loop. LangChain runs each agent turn.**

The Temporal workflow owns the session: the conversation, every tool result, policy, human approvals, timers, and calls into other services. LangChain does the work inside each model call: it formats messages, binds tools, calls the model, and adapts MCP tools. When anything fails, Temporal retries the step or rebuilds the session from its event history.

This is a demo for learning, not production code. Everything runs on your machine, and the incident data, ticketing, paging, monitoring, and CMDB are simulated.

What you'll see
Quick start
Prerequisites
How it works
Using the console
Scenarios
Running options, including choosing an LLM and your own Temporal service
Configuration
Command reference
Tests
Troubleshooting
Project layout
Security notes
Further reading

What you'll see

Capability	Where it shows up
LangChain inside Temporal activities	Every model decision is a `call_llm` activity using a LangChain chat model (`ChatAnthropic` by default) with bound tools
Bring your own LLM	Claude by default; OpenAI, Azure OpenAI, Bedrock, Gemini, Ollama, or any OpenAI-compatible server through configuration
MCP integration	The agent discovers tools from a separate MCP server at run time, t