reworkd / agentgpt Goto Github PK

Task Execution Response:
I apologize, but as an AI language model, I am not capable of physically retrieving bread and condiments from the pantry and fridge. I suggest you manually retrieve the required items and proceed with making your sandwich. However, I can assist you with providing recipe suggestions or tips on how to make a delicious sandwich if you'd like.

Full log:
Embarking on a new goal:

make me sandwitch

Thinking...(Restart if this takes more than 30 seconds)
Added task:

Retrieve bread and condiments from pantry and fridge

Added task:

Retrieve desired sandwich fillings from fridge or pantry

Added task:

Assemble sandwich according to preferred order and cut in half

Thinking...(Restart if this takes more than 30 seconds)
Executing "Retrieve bread and condiments from pantry and fridge"

Task Execution Response: I apologize, but as an AI language model, I am not capable of physically retrieving bread and condiments from the pantry and fridge. I suggest you manually retrieve the required items and proceed with making your sandwich. However, I can assist you with providing recipe suggestions or tips on how to make a delicious sandwich if you'd like.

Thinking...(Restart if this takes more than 30 seconds)
Added task:

Retrieve desired sandwich fillings from fridge or pantry

Added task:

Assemble sandwich according to preferred order and cut in half

Thinking...(Restart if this takes more than 30 seconds)
Executing "Retrieve desired sandwich fillings from fridge or pantry"

Task Execution Response:

I'm sorry, but I cannot retrieve the desired sandwich fillings from the fridge or pantry as I do not have access to physical objects. However, I can suggest some common sandwich fillings such as turkey, ham, cheese, lettuce, tomato, mayo, mustard, and avocado. Would you like me to provide a recipe for a specific type of sandwich?

Thinking...(Restart if this takes more than 30 seconds)
Added task:

Provide a recipe for a specific type of sandwich based on preferred fillings and ingredients.

Thinking...(Restart if this takes more than 30 seconds)
Executing "Assemble sandwich according to preferred order and cut in half"

I'm sorry, but as an AI language model, I'm not capable of physically assembling and cutting a sandwich for you. However, I can provide you with instructions on how to make a sandwich. Here is a simple recipe for a classic sandwich:

Ingredients:

2 slices of bread
Your preferred sandwich fillings (e.g. ham, cheese, lettuce, tomato, mayo, mustard)

Instructions:

Gather all the ingredients and place them on a clean surface.
Take one slice of bread and spread your preferred sauce or condiment on it (e.g. mayo, mustard).
Add your sandwich fillings on top of the sauce. You can add as much or as little as you like.
Place the second slice of bread on top of the fillings.
Use a sharp knife to cut the sandwich in half.
Enjoy your freshly made sandwich!

I hope this helps you make a delicious sandwich!

Thinking...(Restart if this takes more than 30 seconds)
Added task:

Retrieve desired sandwich fillings from fridge or pantry

Added task:

Provide a recipe for a specific type of sandwich based on preferred fillings and ingredients.

We're sorry, because this is a demo, we cannot have our agents running for too long. Note, if you desire longer runs, please provide your own API key in Settings. Shutting down.

Would be great to run this on Replit mobile app

The following output from the agent looks difficult to read the code

error - RangeError [ERR_HTTP_INVALID_STATUS_CODE]: Invalid status code: 0
at new NodeError (node:internal/errors:399:5)
at ServerResponse.writeHead (node:_http_server:344:11)
at ServerResponse.writeHead (E:\otherfile\AgentGPT\node_modules\next\dist\compiled\compression\index.js:46:263)
at ServerResponse._implicitHeader (node:_http_server:335:8)
at ServerResponse.end (E:\otherfile\AgentGPT\node_modules\next\dist\compiled\compression\index.js:22:749)
at DevServer.runEdgeFunction (E:\otherfile\AgentGPT\node_modules\next\dist\server\next-server.js:1461:41)
at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
at async DevServer.runApi (E:\otherfile\AgentGPT\node_modules\next\dist\server\next-server.js:468:47)
at async Object.fn (E:\otherfile\AgentGPT\node_modules\next\dist\server\next-server.js:751:37)
at async Router.execute (E:\otherfile\AgentGPT\node_modules\next\dist\server\router.js:253:36)
at async DevServer.run (E:\otherfile\AgentGPT\node_modules\next\dist\server\base-server.js:384:29)
at async DevServer.run (E:\otherfile\AgentGPT\node_modules\next\dist\server\dev\next-dev-server.js:743:20)
at async DevServer.handleRequest (E:\otherfile\AgentGPT\node_modules\next\dist\server\base-server.js:322:20) {
code: 'ERR_HTTP_INVALID_STATUS_CODE'

The styling may cause text or buttons to appear off screen with some resolutions. We should also improve the design to take up as much space as possible in smaller screens. We should also reduce padding in buttons / input fields in smaller screens. If you work on this, please include an image of the result.

npx db push
npm ERR! could not determine executable to run

npm ERR! A complete log of this run can be found in: /home/chris/.npm/_logs/2023-04-10T07_34_34_100Z-debug-0.log

cat ~/.npm/_logs/2023-04-10T07_34_34_100Z-debug-0.log
0 verbose cli /usr/bin/node /usr/local/lib/node_modules/npm/bin/npm-cli.js
1 info using [email protected]
2 info using [email protected]
3 timing npm:load:whichnode Completed in 2ms
4 timing config:load:defaults Completed in 2ms
5 timing config:load:file:/usr/local/lib/node_modules/npm/npmrc Completed in 1ms
6 timing config:load:builtin Completed in 1ms
7 timing config:load:cli Completed in 1ms
8 timing config:load:env Completed in 0ms
9 timing config:load:file:/home/chris/ai/data/aiAgent/AgentGPT/.npmrc Completed in 0ms
10 timing config:load:project Completed in 3ms
11 timing config:load:file:/home/chris/.npmrc Completed in 0ms
12 timing config:load:user Completed in 0ms
13 timing config:load:file:/etc/npmrc Completed in 0ms
14 timing config:load:global Completed in 0ms
15 timing config:load:setEnvs Completed in 1ms
16 timing config:load Completed in 9ms
17 timing npm:load:configload Completed in 9ms
18 timing npm:load:mkdirpcache Completed in 0ms
19 timing npm:load:mkdirplogs Completed in 0ms
20 verbose title npm exec db push
21 verbose argv "exec" "--" "db" "push"
22 timing npm:load:setTitle Completed in 1ms
23 timing config:load:flatten Completed in 2ms
24 timing npm:load:display Completed in 2ms
25 verbose logfile logs-max:10 dir:/home/chris/.npm/_logs/2023-04-10T07_34_34_100Z-
26 verbose logfile /home/chris/.npm/_logs/2023-04-10T07_34_34_100Z-debug-0.log
27 timing npm:load:logFile Completed in 5ms
28 timing npm:load:timers Completed in 0ms
29 timing npm:load:configScope Completed in 0ms
30 timing npm:load Completed in 19ms
31 silly logfile start cleaning logs, removing 1 files
32 silly logfile done cleaning log files
33 timing arborist:ctor Completed in 0ms
34 http fetch GET 200 https://registry.npmjs.org/db 172ms (cache revalidated)
35 timing command:exec Completed in 194ms
36 verbose stack Error: could not determine executable to run
36 verbose stack at getBinFromManifest (/usr/local/lib/node_modules/npm/node_modules/libnpmexec/lib/get-bin-from-manifest.js:17:23)
36 verbose stack at exec (/usr/local/lib/node_modules/npm/node_modules/libnpmexec/lib/index.js:185:15)
36 verbose stack at async module.exports (/usr/local/lib/node_modules/npm/lib/cli.js:89:5)
37 verbose pkgid [email protected]
38 verbose cwd /home/chris/ai/data/aiAgent/AgentGPT
39 verbose Linux 5.19.0-38-generic
40 verbose node v18.15.0
41 verbose npm v9.6.4
42 error could not determine executable to run
43 verbose exit 1
44 timing npm Completed in 224ms
45 verbose code 1
46 error A complete log of this run can be found in: /home/chris/.npm/_logs/2023-04-10T07_34_34_100Z-debug-0.log

I'm not familar with npx so ?

I'm currently using a ChatGPT client* which is completely functional in the browser, and the users OpenAI key is only stored in the browser's local-storage database. This setup allows other people to use their own key with the AI tool.

It would be great to see this project work in that mode too. I understand there could be CORS issues, but sometimes all you need for that is a JSON-proxy. Thanks for this interesting project!

Hello,
I have test the web version of Agent GPT and i provide my own API key and it start to work but after sometimes it cut and saying we are sorry, because this a demo we cannot .......
Please i would like to know what is the issue

This is a big undertaking. As a user, I'd like to be able to use my own local models to not be crippled by limitations of OpenAI models. The opensource model scene is getting huge and we want to support this as best as possible. The issue is, given we are a product first and foremost, connection to local models must work in both prod and dev and must be as simple as possible to setup.

One option is WindowAI: https://twitter.com/xanderatallah/status/1643356106670981122. This is an extension that will allow a site to connect to a model locally.

Before implementing anything, we should first investigate possibilities and report back in this issue.

We need to integrate with langchang agents api to allow the agent to make google searches for answers. This will need a google api key that is production allowlisted

Currently, we have the option of saving the run as an image. This is good for sharability but not effective in actually using output from the tool. The save button should open up a dialog that allows the user to either download the image, download a pdf/text document, or save the conversation to the clipboard

I installed the version from the main branch, commit 06e781e

I asked for some HTML with some content (not important), but the HTML being generated is truncated.

The truncated part is not even that big. The first test the agent output ~400 bytes. In other test, ~800 bytes, always truncated.

Is there a size limit for the output? Any way to increase the size limit?

Environment variables loaded from .env
Prisma schema loaded from prisma/schema.prisma

✔ Generated Prisma Client (4.12.0 | library) to ./node_modules/@prisma/client in 177ms
You can now start using Prisma Client in your code. Reference: https://pris.ly/d/client

import { PrismaClient } from '@prisma/client'
const prisma = new PrismaClient()

Environment variables loaded from .env
Prisma schema loaded from prisma/schema.prisma
Datasource "db": SQLite database "db.sqlite" at "file:../db/db.sqlite"

**Error: Prisma Migrate has detected that the environment is non-interactive, which is not supported.

prisma migrate dev is an interactive command designed to create new migrations and evolve the database in development.
To apply existing migrations in deployments, use prisma migrate deploy.
See https://www.prisma.io/docs/reference/api-reference/command-reference#migrate-deploy**

The command '/bin/sh -c npx prisma generate && npx prisma migrate dev --name init --preview-feature && npx prisma db push' returned a non-zero code: 1

It looks like there is a problem with "prisma migrate dev", I tried to use "prisma migrate deploy" to replace it directly and it seems to be a problem

**Error: P3005

The database schema is not empty. Read more about how to baseline an existing production database: https://pris.ly/d/migrate-baseline**

The readme instructions should also include the instruction to cd AgentGPT

I would suggest adding a license to this repo.

Hello,

I would love to run AgentGPT with local models using https://github.com/abetlen/llama-cpp-python which provides an emulated openai api.

Can the API host be made user settable via a
OPENAI_API_HOST env variable?

The number of iterations should be configurable. How can I change it manually (At least it would be 100 Loops)?

As a user, I want to be able to set goals like: "Make a book of art" and have AgentGPT output images.

To do this, we should use LangChain agents and connect to an external API to generate images. We may consider making a chain of a model coming up with a prompt, then making the call out.

Hello, thanks for this awesome work.

If we discuss this topic regarding differences, pros & cons with https://github.com/Torantulino/Auto-GPT I am open to craft a PR to the docs.

As a user, I may not want to be tied down to OpenAI. There are a lot of companies offering APIs for LLMs. We should explore what are the best, and provide an interface to select between them

The agent doesn't stop when it should (it has completed and no new tasks added)

The button copies the conversation, but to let the users know it is successfully copied, the copy button should change to coped.

You can clearly see the home page is overflowing in x and y directions and a part of logo is not visible.
Screen Size: 15.6 inches.

When the agent is done , can it give a downloadable text file or pdf with the data from the agent that contains it's findings.

npx prisma db push

Prisma schema loaded from prisma/schema.prisma
Datasource "db": PostgreSQL database

Error: Prisma schema validation - (get-config wasm)
Error code: P1012
error: Environment variable not found: DATABASE_URL.
  -->  schema.prisma:7
   |
 6 |     provider = "postgresql"
 7 |     url      = env("DATABASE_URL")
   |

Validation Error Count: 1
[Context: getConfig]

Prisma CLI Version : 4.12.0

M1 Mac OS

Azure api-key is different from openai api-key. And it should be used.

Instead of needing to wait for approval, automatically test + deploy with GA (unless it's too expensive on vercel idk those details)

https://vercel.com/guides/how-can-i-use-github-actions-with-vercel

As a user, I want the ability to copy an individual message I found interesting.

How this should work is on hover, a copy icon should appear on the bottom right of the message. Clicking anywhere on the message should then copy the message in its entirety into the clip board. There should be some indication that the message was copied

Perhaps such actions should call for review from the user at this stage.

It would be great to empower AgentGPT to actually take action beyond gathering data.

• A list of emails, SMS, tweets, to be sent out
• A document, spreadsheet, photo, or video that can be developed and made ready for use
• A instance of a server (wordpress, BTCPayServer, SMTP, etc) fully configured with a host like Digital Ocean or AWS.
• An invoice that can be paid via Wise or bitcoin

Having these actions "suggested" in a list so they may be approved with the push of a button would save time. Currently, it seems any actions have to be completely done manually by the user.

If an agent gives me 500 great tasks to carry out manually, that's still more like an advisor than an autonomous agent.

The Limited Settings limit the capabilities of the AI. To improve this, I suggest adding additional settings, especially when combined with a custom key provided by the user. These settings could include:

• Model
• Loop Limit
• Temperature

These settings can be changed locally, but this would also be beneficial for the web UI in general.

Right now, it is unclear where my API key is going after I copy and paste it into the page. Is it stored in browser? Deleted right away? This should be front and center in the settings page.

Getting this error when "deploy agent" button is clicked (sensible data replaced with ... )
The prompt was a simple "plan a week of ketogenic diet"
On the frontend side I get "ERROR retrieving initial tasks array. Make sure your API key is not the free tier, make your goal more clear, or revise your goal such that it is within our model's policies to run. Shutting Down."
Is there a free tier openai API key? I'm not aware of, anyway mine under a billing usage

error - AxiosError: Request failed with status code 401 at createError (file:///.../AgentGPT/node_modules/langchain/dist/util/axios-fetch-adapter.js:307:16) at settle (file:///.../AgentGPT/node_modules/langchain/dist/util/axios-fetch-adapter.js:26:16) at file:///.../AgentGPT/node_modules/langchain/dist/util/axios-fetch-adapter.js:176:19 at new Promise (<anonymous>) at fetchAdapter (file:///.../AgentGPT/node_modules/langchain/dist/util/axios-fetch-adapter.js:168:12) at process.processTicksAndRejections (node:internal/process/task_queues:95:5) { config: { transitional: { silentJSONParsing: true, forcedJSONParsing: true, clarifyTimeoutError: false }, adapter: [AsyncFunction: fetchAdapter], transformRequest: [ [Function: transformRequest] ], transformResponse: [ [Function: transformResponse] ], timeout: 0, xsrfCookieName: 'XSRF-TOKEN', xsrfHeaderName: 'X-XSRF-TOKEN', maxContentLength: -1, maxBodyLength: -1, validateStatus: [Function: validateStatus], headers: { Accept: 'application/json, text/plain, */*', 'Content-Type': 'application/json', 'User-Agent': 'OpenAI/NodeJS/3.2.1', Authorization: 'Bearer ...' }, method: 'post', data: '{"model":"gpt-3.5-turbo","temperature":0.9,"top_p":1,"frequency_penalty":0,"presence_penalty":0,"n":1,"max_tokens":300,"stream":false,"messages":[{"role":"user","content":"You are an autonomous task creation AI called AgentGPT. You have the following objective plan a week of ketogenic diet. Create a list of zero to three tasks to be completed by your AI system such that your goal is more closely reached or completely reached. Return the response as an array of strings that can be used in JSON.parse()"}]}', url: 'https://api.openai.com/v1/chat/completions' }, request: Request { [Symbol(realm)]: { settingsObject: [Object] }, [Symbol(state)]: { method: 'POST', localURLsOnly: false, unsafeRequest: false, body: [Object], client: [Object], reservedClient: null, replacesClientId: '', window: 'client', keepalive: false, serviceWorkers: 'all', initiator: '', destination: '', priority: null, origin: 'client', policyContainer: 'client', referrer: 'client', referrerPolicy: '', mode: 'cors', useCORSPreflightFlag: false, credentials: 'same-origin', useCredentials: false, cache: 'default', redirect: 'follow', integrity: '', cryptoGraphicsNonceMetadata: '', parserMetadata: '', reloadNavigation: false, historyNavigation: false, userActivation: false, taintedOrigin: false, redirectCount: 0, responseTainting: 'basic', preventNoCacheCacheControlHeaderModification: false, done: false, timingAllowFailed: false, headersList: [HeadersList], urlList: [Array], url: [URL] }, [Symbol(signal)]: AbortSignal { aborted: false }, [Symbol(headers)]: HeadersList { [Symbol(headers map)]: [Map], [Symbol(headers map sorted)]: null } }, response: { ok: false, status: 401, statusText: 'Unauthorized', headers: HeadersList { [Symbol(headers map)]: [Map], [Symbol(headers map sorted)]: null }, config: { transitional: [Object], adapter: [AsyncFunction: fetchAdapter], transformRequest: [Array], transformResponse: [Array], timeout: 0, xsrfCookieName: 'XSRF-TOKEN', xsrfHeaderName: 'X-XSRF-TOKEN', maxContentLength: -1, maxBodyLength: -1, validateStatus: [Function: validateStatus], headers: [Object], method: 'post', data: '{"model":"gpt-3.5-turbo","temperature":0.9,"top_p":1,"frequency_penalty":0,"presence_penalty":0,"n":1,"max_tokens":300,"stream":false,"messages":[{"role":"user","content":"You are an autonomous task creation AI called AgentGPT. You have the following objective plan a week of ketogenic diet. Create a list of zero to three tasks to be completed by your AI system such that your goal is more closely reached or completely reached. Return the response as an array of strings that can be used in JSON.parse()"}]}', url: 'https://api.openai.com/v1/chat/completions' }, request: Request { [Symbol(realm)]: [Object], [Symbol(state)]: [Object], [Symbol(signal)]: [AbortSignal], [Symbol(headers)]: [HeadersList] }, data: { error: [Object] } }, page: '/api/chain' }

There are OpenAI API compatible endpoints for other models, such as Basaran: https://github.com/hyperonym/basaran

Adding an endpoint variable to the .env would allow users to use any LLM with AgentGPT, not just OpenAI's endpoint.

This is hugely useful for a number of reasons. But the biggest one is that local models don't rack up an API bill.

Another use case for this is proxy endpoints which log the requests/responses. Lots of people use OpenAI's models with endpoint proxies (myself included).

The tool is very powerful, but the amount of information generated after the program finishes execution is huge. The current export methods are difficult to organize. Is it possible to use a tree structure to organize the results by levels, for example, exporting in formats such as markdown or JSON?

Name: DateGPT
Goal: Give the current date
Logs:

Embarking on a new goal:Give the current date
Thinking...(Restart if this takes more than 30 seconds)
Added task:Retrieve the current date from system clock
Added task:Format date as a string
Added task:Return date string as output
Thinking...(Restart if this takes more than 30 seconds)
Executing "Retrieve the current date from system clock"Executing task: Retrieve the current date from system clock Response: Today's date is [insert current date here].
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Format date as a string
Added task:Execute task: Return date string as output
Thinking...(Restart if this takes more than 30 seconds)
Executing "Format date as a string"The current date is September 28, 2021.
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Return date string as output
Thinking...(Restart if this takes more than 30 seconds)
Executing "Return date string as output"Today's date is October 20, 2021.
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Format date as a string
Thinking...(Restart if this takes more than 30 seconds)
Executing "Execute task: Format date as a string"Executing task: Format date as a string. Today's date is: July 28, 2021. Returning response as a string: "July 28, 2021".
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Return date and time string as output
Thinking...(Restart if this takes more than 30 seconds)
Executing "Execute task: Return date string as output"Sure, the current date is [insert current date here].
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Format date as a string
Thinking...(Restart if this takes more than 30 seconds)
Executing "Execute task: Return date string as output"As an AI language model, I do not have access to real-time data or system clock. However, if I am to provide a sample output, it would be something like: "Today is August 25, 2021."
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Format date and time as a string
Thinking...(Restart if this takes more than 30 seconds)
Executing "Execute task: Format date as a string"Executing task: Format date as a string. The current date is: August 3, 2021. String formatted date: 08/03/2021 Returning formatted date as a string.
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Return date and time string as output
Added task:Execute task: Format date and time as a string
Thinking...(Restart if this takes more than 30 seconds)
Executing "Execute task: Return date and time string as output"Sure, I can help with that! The current date and time is [insert current date and time here].
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Format date as a string
Added task:Execute task: Format date and time as a string
Thinking...(Restart if this takes more than 30 seconds)
Executing "Execute task: Format date as a string"Task response: Today's date is: 2021-10-17
Thinking...(Restart if this takes more than 30 seconds)
Added task:Execute task: Format date and time as a string
Added task:Execute task: Return date and time string as output
Added task:Execute task: Format date and time as a string
Added task:Execute task: Format date and time as a string
We're sorry, because this is a demo, we cannot have our agents running for too long. Shutting down.

This looks pretty stupid, especially the "Execute task: Format date as a string" part that seems the AI got stuck on.

I am aware that this is a beta version, however the website title seems off:

Assign me this issue, I will fix it and we will be done with these types of trivial issues in no time..

How to keep the program running continuously on VPS system CentOS 7.9? I cannot access the web page after disconnecting the SSH connection. Thank you very much.

Many of your users have access to GPT-4 already. I checked the settings but there is no option to use it. I checked the code, and the model name is baked-in as gpt-3.5-turbo. It would be very nice to have a GPT-4 option considering the difference in quality.
(Might be worth adding a notice that using GPT-4 will be slower and more expensive.)

Where do I put my Open AI API Key or other AI Key?

It seems like there's a mismatch for the database URL in the prism schema:

`> npx prisma db push
Environment variables loaded from .env
Prisma schema loaded from prisma/schema.prisma
Datasource "db": PostgreSQL database

Error: Prisma schema validation - (get-config wasm)
Error code: P1012
error: Error validating datasource db: the URL must start with the protocol postgresql:// or postgres://.
--> schema.prisma:7
|
6 | provider = "postgresql"
7 | url = env("DATABASE_URL")
|

Validation Error Count: 1
[Context: getConfig]`

I am in china, so I need to use vpn to get the response from openai, but when the vpn is on, the error occurs.
error - RangeError [ERR_HTTP_INVALID_STATUS_CODE]: Invalid status code: 0
at new NodeError (node:internal/errors:399:5)
at ServerResponse.writeHead (node:_http_server:344:11)
at ServerResponse.writeHead (E:\AgentGPT-main\node_modules\next\dist\compiled\compression\index.js:46:263)
at ServerResponse._implicitHeader (node:_http_server:335:8)
at ServerResponse.end (E:\AgentGPT-main\node_modules\next\dist\compiled\compression\index.js:22:749)
at DevServer.runEdgeFunction (E:\AgentGPT-main\node_modules\next\dist\server\next-server.js:1461:41)
at process.processTicksAndRejections (node:internal/process/task_queues:95:5)
at async DevServer.runApi (E:\AgentGPT-main\node_modules\next\dist\server\next-server.js:468:47)
at async Object.fn (E:\AgentGPT-main\node_modules\next\dist\server\next-server.js:751:37)
at async Router.execute (E:\AgentGPT-main\node_modules\next\dist\server\router.js:253:36)
at async DevServer.run (E:\AgentGPT-main\node_modules\next\dist\server\base-server.js:384:29)
at async DevServer.run (E:\AgentGPT-main\node_modules\next\dist\server\dev\next-dev-server.js:743:20)
at async DevServer.handleRequest (E:\AgentGPT-main\node_modules\next\dist\server\base-server.js:322:20) {
code: 'ERR_HTTP_INVALID_STATUS_CODE'

Hello, I am running on localhost but it keeps asking for my API key. Is that expected behaviour?

error - RangeError [ERR_HTTP_INVALID_STATUS_CODE]: Invalid status code: 0 at new NodeError (node:internal/errors:399:5) at ServerResponse.writeHead (node:_http_server:344:11)

What is the problem?

• Stops at the 3rd thinking step

elements that receive a click and navigate to a url should properly be anchor elements

<DrawerItem
  icon={<FaGithub />}
  text="GitHub"
  onClick={() =>
    window.open("https://github.com/reworkd/AgentGPT", "_blank")
  }
/>

should really be:

<a href="https://github.com/reworkd/AgentGPT" target="_blank">GitHub</a>