First, thank you for taking the time to be here—we really appreciate it. ⭐️ We're running a one-week beta testing program for potpie with engineers from 25 companies to get early feedback and polish the product before its public launch. Given your experience as an engineer, your participation in the beta program will help us shape the product in a more customer-centric way..

This is a one-pager on everything you need to start testing the product.

Here are some pointers for evaluating the product, though they are not exhaustive:

✅ Usefulness: Does it solve a problem for you?

✅ Developer experience: Is it easy to use?

✅ Time to value: how fast were you able to realize that this tool is valuable

✅ Accuracy and performance: How satisfactory was the agent’s performance

🐞 Bugs

Additionally if you encounter bugs, please report it on our discord channel #ifoundabug

How to make potpie useful?

We currently have built 6 agents on the platform and here are details and sample use cases for each of them

<aside> 🤖

Codebase Q&A agent

This agent specializes in answering questions about the codebase using the knowledge graph and code analysis tools. You can type @ followed by function/file name to provide entry point to the agent if you want to ask it about something specific.

Use it for asking questions like

✅  Explain how a document is inserted into the database?

✅  Explain how authorisation is implemented in the codebase.

</aside>

<aside> 🤖

Unit test agent

This agent takes a reference of a function or class name as innput and helps in generating test plan and unit tests for it. You can play and edit tests and plans created by it. Type @ followed by function name to provide the function reference to the agent.

Use it for asking questions like

✅  Help me plan and write a unit test for @create_document

✅  Add an additional test to assert that the value returned is not null

</aside>

<aside> 🤖

Integration test agent

This agent specializes in generating integration test plan and test code for code snippets from the knowledge graph based on given function names of entry points. It takes the entry point reference as input and understands the flow. Works best with Python, JavaScript, TypeScript. Type @ followed by function name to provide entry point to the agent.

Use it for asking questions like

✅ Generate a test plan for the POST /documents/{collection_name} API at @create_document

✅ This test fails to mock the MongoService class method, update the mocked path in all tests to follow the format path.to.file.Class.function

</aside>

<aside> 🤖

Debugging with Knowledge Graph agent

This agent specialized in analysing stacktraces and errors and relate it to your codebase. It will help you by providing debugging directions, add print debugging statements for you and help you come to the root cause iteratively.

Use it for asking questions like

✅  I’m getting a 401 unauthorized error from @update_document API , help me debug

✅  Why am I getting this error. Stacktrace :

Traceback (most recent call last):
File "/app/training_unit.py", line 125, in delete_training_unit
raise HTTPException(
fastapi.exceptions.HTTPException: Could not delete training unit. Error: (sqlite3.DatabaseError) Failed to delete record

✅ Help me fix the TypeError in this line query.lower().split() of @query_vector_store

</aside>

<aside> 🤖

Code Changes Agent

This agent specializes in generating detailed analysis of code changes in your current branch compared to default branch. Works best with Py, JS, TS

Use it for asking questions like

✅ Which APIs are impacted by my current change?

✅ Help me under any cascading effects of my changes.

✅ Review my changes and suggest improvements

</aside>