MCPcopy Index your code
hub / github.com/Hendrixer/agents-production

github.com/Hendrixer/agents-production @main

Chat with this repo
repository ↗ · DeepWiki ↗ · + Follow
43 symbols 132 edges 29 files 0 documented · 0%
What it actually does AI analysis from the code graph — generated when you open this
loading…
README

Frontend Masters

Prototype to Production: Advanced AI Apps

Setup Instructions

This repo requires Node.js version 20+ or bun v1.0.20.

The main branch contains the final application. To code along with the workshop, checkout the step/1 branch. You will also need an API Key from OpenAI.

git clone https://github.com/Hendrixer/agents-production.git
cd agents-production
git checkout step/1
npm install # or bun install

To run the project:

npm start
# or
bun run index.ts

OpenAI API Key

Create an API Key from OpenAI and save it in a .env file:

OPENAI_API_KEY='YOUR_API_KEY'

image

OpenAI requires you to add a minimum of $5 USD credit to your account. Go to the billing page and add credits.

image

[!NOTE]
It can take up to 20 minutes for the credits to be available via the API. So if you get an insufficient_quota error, try again in a few minutes.

Upstash Account

A free Upstash account is required for hosting a serverless vector database. The setup is demonstrated in the course. You can find the details in the course notes

Windows Environments

The dynamic imports may be incompatible depending on the windows environment, so you may need to run the evals directly. Add the environment import to the top of each eval:

import 'dotenv/config'

Then run the eval with the full path to the file. For example:

npx tsx evals/experiments/reddit.eval.ts

Extension points exported contracts — how you extend this code

ToolFn (Interface)
(no doc)
types.ts
ExperimentGraphProps (Interface)
(no doc)
dashboard/src/components/ExperimentGraph.tsx
Score (Interface)
(no doc)
types.ts
Run (Interface)
(no doc)
types.ts
Set (Interface)
(no doc)
types.ts
Experiment (Interface)
(no doc)
types.ts

Core symbols most depended-on inside this repo

runLLM
called by 7
src/llm.ts
getMessages
called by 4
src/memory.ts
runEval
called by 4
evals/evalTools.ts
createToolCallMessage
called by 4
evals/experiments/allTools.eval.ts
runTool
called by 3
src/toolRunner.ts
getDb
called by 3
src/memory.ts
addMessages
called by 3
src/memory.ts
saveToolResponse
called by 3
src/memory.ts

Shape

Function 36
Interface 7

Languages

TypeScript100%

Modules by API surface

src/memory.ts7 symbols
types.ts6 symbols
evals/evalTools.ts5 symbols
src/llm.ts3 symbols
src/agent.ts3 symbols
src/ui.ts2 symbols
dashboard/src/components/ExperimentGraph.tsx2 symbols
src/tools/reddit.ts1 symbols
src/tools/movieSearch.ts1 symbols
src/tools/generateImage.ts1 symbols
src/tools/dadJoke.ts1 symbols
src/toolRunner.ts1 symbols

For agents

$ claude mcp add agents-production \
  -- python -m otcore.mcp_server <graph>

⬇ download graph artifact