Skip to content

Latest commit

 

History

History
252 lines (178 loc) · 13.7 KB

readme.md

File metadata and controls

252 lines (178 loc) · 13.7 KB

🐣 smol developer

Deploy agent on e2b button

Human-centric & Coherent Whole Program Synthesis aka your own personal junior developer

Build the thing that builds the thing! a smol dev for every dev in every situation

This is a "junior developer" agent (aka smol dev) that either:

  1. scaffolds an entire codebase out for you once you give it a product spec
  2. gives you basic building blocks to have a smol developer inside of your own app.

Instead of making and maintaining specific, rigid, one-shot starters, like create-react-app, or create-nextjs-app, this is basically is or helps you make create-anything-app where you develop your scaffolding prompt in a tight loop with your smol dev.

After the successful initial v0 launch, smol developer was rewritten to be even smol-ler, and importable from a library!

Basic Usage

In Git Repo mode

# install
git clone https://github.com/smol-ai/developer.git
cd developer
poetry install # install dependencies. pip install poetry if you need

# run
python main.py "a HTML/JS/CSS Tic Tac Toe Game" # defaults to gpt-4-0613
# python main.py "a HTML/JS/CSS Tic Tac Toe Game" --model=gpt-3.5-turbo-0613

# other cli flags
python main.py --prompt prompt.md # for longer prompts, move them into a markdown file
python main.py --prompt prompt.md --debug True # for debugging
This lets you develop apps as a human in the loop, as per the original version of smol developer.

engineering with prompts, rather than prompt engineering

The demo example in prompt.md shows the potential of AI-enabled, but still firmly human developer centric, workflow:

  • Human writes a basic prompt for the app they want to build
  • main.py generates code
  • Human runs/reads the code
  • Human can:
    • simply add to the prompt as they discover underspecified parts of the prompt
    • manually runs the code and identifies errors
    • paste the error into the prompt just like they would file a GitHub issue
    • for extra help, they can use debugger.py which reads the whole codebase to make specific code change suggestions

Loop until happiness is attained. Notice that AI is only used as long as it is adding value - once it gets in your way, just take over the codebase from your smol junior developer with no fuss and no hurt feelings. (we could also have smol-dev take over an existing codebase and bootstrap its own prompt... but that's a Future Direction)

In this way you can use your clone of this repo itself to prototype/develop your app.

In Library mode

This is the new thing in smol developer v1! Add smol developer to your own projects!

pip install smol_dev

Here you can basically look at the contents of main.py as our "documentation" of how you can use these functions and prompts in your own app:

from smol_dev.prompts import plan, specify_file_paths, generate_code_sync

prompt = "a HTML/JS/CSS Tic Tac Toe Game"

shared_deps = plan(prompt) # returns a long string representing the coding plan

# do something with the shared_deps plan if you wish, for example ask for user confirmation/edits and iterate in a loop

file_paths = specify_file_paths(prompt, shared_deps) # returns an array of strings representing the filenames it needs to write based on your prompt and shared_deps. Relies on OpenAI's new Function Calling API to guarantee JSON.

# do something with the filepaths if you wish, for example display a plan

# loop through file_paths array and generate code for each file
for file_path in file_paths:
    code = generate_code_sync(prompt, shared_deps, file_path) # generates the source code of each file

    # do something with the source code of the file, eg. write to disk or display in UI
    # there is also an async `generate_code()` version of this

In API mode (via Agent Protocol)

To start the server run:

poetry run api

or

python smol_dev/api.py

and then you can call the API using either the following commands:

To create a task run:

curl --request POST \
  --url http://localhost:8000/agent/tasks \
  --header 'Content-Type: application/json' \
  --data '{
	"input": "Write simple script in Python. It should write '\''Hello world!'\'' to hi.txt"
}'

You will get a response like this:

{"input":"Write simple script in Python. It should write 'Hello world!' to hi.txt","task_id":"d2c4e543-ae08-4a97-9ac5-5f9a4459cb19","artifacts":[]}

Then to execute one step of the task copy the task_id you got from the previous request and run:

curl --request POST \
  --url http://localhost:8000/agent/tasks/<task-id>/steps

or you can use Python client library:

from agent_protocol_client import AgentApi, ApiClient, TaskRequestBody

...

prompt = "Write simple script in Python. It should write 'Hello world!' to hi.txt"

async with ApiClient() as api_client:
    # Create an instance of the API class
    api_instance = AgentApi(api_client)
    task_request_body = TaskRequestBody(input=prompt)

    task = await api_instance.create_agent_task(
        task_request_body=task_request_body
    )
    task_id = task.task_id
    response = await api_instance.execute_agent_task_step(task_id=task_id)

...

examples/prompt gallery

I'm actively seeking more examples, please PR yours!

sorry for the lack of examples, I know that is frustrating but I wasnt ready for so many of you lol

major forks/alternatives

please send in alternative implementations, and deploy strategies on alternative stacks!

innovations and insights

Please subscribe to https://latent.space/ for a fuller writeup and insights and reflections

  • Markdown is all you need - Markdown is the perfect way to prompt for whole program synthesis because it is easy to mix english and code (whether variable_names or entire ``` code fenced code samples)
    • turns out you can specify prompts in code in prompts and gpt4 obeys that to the letter
  • Copy and paste programming
    • teaching the program to understand how to code around a new API (Anthropic's API is after GPT3's knowledge cutoff) by just pasting in the curl input and output
    • pasting error messages into the prompt and vaguely telling the program how you'd like it handled. it kind of feels like "logbook driven programming".
  • Debugging by cating the whole codebase with your error message and getting specific fix suggestions - particularly delightful!
  • Tricks for whole program coherence - our chosen example usecase, Chrome extensions, have a lot of indirect dependencies across files. Any hallucination of cross dependencies causes the whole program to error.
    • We solved this by adding an intermediate step asking GPT to think through shared_dependencies.md, and then insisting on using that in generating each file. This basically means GPT is able to talk to itself...
    • ... but it's not perfect, yet. shared_dependencies.md is sometimes not comperehensive in understanding what are hard dependencies between files. So we just solved it by specifying a specific name in the prompt. felt dirty at first but it works, and really it's just clear unambiguous communication at the end of the day.
    • see prompt.md for SOTA smol-dev prompting
  • Low activation energy for unfamiliar APIs
    • we have never really learned css animations, but now can just say we want a "juicy css animated red and white candy stripe loading indicator" and it does the thing.
    • ditto for Chrome Extension Manifest v3 - the docs are an abject mess, but fortunately we don't have to read them now to just get a basic thing done
    • the Anthropic docs (bad bad) were missing guidance on what return signature they have. so just curl it and dump it in the prompt lol.
  • Modal is all you need - we chose Modal to solve 4 things:
    • solve python dependency hell in dev and prod
    • parallelizable code generation
    • simple upgrade path from local dev to cloud hosted endpoints (in future)
    • fault tolerant openai api calls with retries/backoff, and attached storage (for future use)

Please subscribe to https://latent.space/ for a fuller writeup and insights and reflections

caveats

We were working on a Chrome Extension, which requires images to be generated, so we added some usecase specific code in there to skip destroying/regenerating them, that we haven't decided how to generalize.

We dont have access to GPT4-32k, but if we did, we'd explore dumping entire API/SDK documentation into context.

The feedback loop is very slow right now (time says about 2-4 mins to generate a program with GPT4, even with parallelization due to Modal (occasionally spiking higher)), but it's a safe bet that it will go down over time (see also "future directions" below).

future directions

things to try/would accept open issue discussions and PRs:

  • specify .md files for each generated file, with further prompts that could finetune the output in each of them
    • so basically like popup.html.md and content_script.js.md and so on
  • bootstrap the prompt.md for existing codebases - write a script to read in a codebase and write a descriptive, bullet pointed prompt that generates it
    • done by smol pm, but its not very good yet - would love for some focused polish/effort until we have quine smol developer that can generate itself lmao
  • ability to install its own dependencies
  • self-heal by running the code itself and use errors as information for reprompting
    • however its a bit hard to get errors from the chrome extension environment so we did not try this
  • using anthropic as the coding layer
    • you can run modal run anthropic.py --prompt prompt.md --outputdir=anthropic to try it
    • but it doesnt work because anthropic doesnt follow instructions to generate file code very well.
  • make agents that autonomously run this code in a loop/watch the prompt file and regenerate code each time, on a new git branch
    • the code could be generated on 5 simultaneous git branches and checking their output would just involve switching git branches