Skip to content

[Model BadCase]: Abrupt stopping before starting to execute a task after planning (agentic use - coding) #37

Open

Description

@ifsin

Basic Information - Models Used

MiniMax-M2

Information about environment and deployment

I'm using the official minimax anthropic API in this terminal coding tool and I have noticed a consistent issue with the M2 model
thinking is working, planning is good but abruptly stops like ".... okay, now I'll start implementing all these changes:"
It continues when I prompt it anything like 'do' or 'go' and starts the tool calling to execute the task.

Description

Steps to reproduce

The bug can be reproduced with the following steps:

use any coding tool with the anthropic API and instruct the model to plan a task and then do it

Expected behavior

abruptly stops after planning

Paste the related screenshots here

Image

Activity

Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Projects

    No projects

      Milestone

      No milestone

      Relationships

      None yet

      Development

      No branches or pull requests

      Issue actions