JetBrains-Research / JetBrains-Research/Python-Solution-Preference
[020] Slack-like Team Chat (MVP)
- Dominant language
- No language data
- Stars
- 2
- Forks
- 0
- PR merge metrics
- No merged PRs in 30d
Description
## Overview
Team chat with 5 fixed users from `assets/users.csv` (columns: `username`, `display_name`, `password`). Login required; invalid credentials rejected.
Build this as a Python backend service with no UI; expose the functionality below through an HTTP API.
## Features
### #general Channel
- Single public channel visible to all users
- Messages returned chronologically (oldest first, newest last)
### Direct Messages (DMs)
- 1:1 private conversations between any two users
- Start DM via user search (case-insensitive partial match on username or display_name, excludes self)
- One DM thread per user pair; selecting same user reopens existing conversation
- A user's DM conversations can be listed alongside `#general`
### Messaging (both #general and DMs)
- Text messages only; whitespace-only rejected, max 2000 characters
- Messages immutable (no edit/delete)
### Search
- Global search across `#general` and the user's own DMs (not others' private DMs)
- Case-insensitive partial text match, results newest-first
- Each result identifies the conversation it belongs to
## Out of Scope
Multiple channels, group DMs, file uploads, reactions, threads, typing indicators
---
**Provided starting files:** see [`020-team-chat/assets/`](https://github.com/JetBrains-Research/Python-Solution-Preference/tree/main/020-team-chat/assets) — these were in the agent's working directory when it solved the task.
Contributor guide
No contributing guide indexed for this repository
Research direction
Start by reading the provided files in 020-team-chat/assets/, especially assets/users.csv and its username, display_name, and password columns. Define the HTTP API around login, #general, one-to-one DMs, messaging, user search, conversation listing, and global search. Done means invalid credentials and invalid messages are rejected, private results stay scoped to the user, and all stated ordering and out-of-scope rules are met.
Written by the indexing model from the issue text.
Assessment
- Tech stack
- python
- Domain
- api, backend
- Issue type
- Feature
- Difficulty
- 5/5
- Estimated time
- Over a week
- Activity status
- Quiet
- Clarity
- Mostly clear
- Newbie friendliness
- 35/100