mirror of
https://github.com/Sea-Haven-Industries/open-swe.git
synced 2026-09-30 13:53:15 +00:00
18 lines
1.3 KiB
Text
18 lines
1.3 KiB
Text
|
---
|
||
|
|
title: 'From the UI'
|
||
|
|
description: 'How to use Open SWE from the UI'
|
||
|
|
---
|
||
|
|
|
||
|
|
(in this page, cover how you can use Open SWE from the UI. Do not include screenshots, I will manually add those later)
|
||
|
|
|
||
|
|
discuss:
|
||
|
|
- Auto mode vs manual mode
|
||
|
|
- chatting with the agent (manager)
|
||
|
|
- what the manager can do: respond to user, create new planning run, send new message to active planning run, resume the planner after interrupted with a response, send new message to active programmer run
|
||
|
|
- what the manager can NOT do: create new programmer run (this can only happen by accepting the proposed plan/auto mode after plan is complete), stop the planner/programmer. re-plan while the programmer is running. open a PR (this will only happen after the programmer finishes. the programmer will always auto-commit after every change though), accept a plan (you need to manually click accept if not running in auto-mode)
|
||
|
|
- overview of how the manager can take actions:
|
||
|
|
- how messages you send to the manager trigger planning runs
|
||
|
|
- how messages you send to the manager can add new messages to planning run while its active
|
||
|
|
- how messages you send to the manager can resume the planner if it's interrupted w/ a proposed plan
|
||
|
|
- how messages you send to the manager can send new messages to active programmer run
|