Repository navigation
[FEATURE]:Pluggable Context Management Method #24893
Description
Feature hasn't been suggested before.
- I have verified this feature I'm about to request hasn't been suggested before.
Describe the enhancement you want to request
When using opencode, the previous model with small context often compresses the context, and sometimes after compression, the AI forgets too much and becomes foolish; Nowadays, using the deepseek-v4-pro model has quite a large context, with 600K contexts used in daily life. It feels like there is a lot of nonsense and wasted tokens in the context. Using default instructions to compress the context is afraid of compressing too much.
At present, context compression has limited options for users. I would like a plugin that allows users to choose which part of the context to compress:
Pluggable context management allows users to use commands to track the usage of the current context and the returned context. LLM needs to classify the context based on its content. Users can choose to uninstall or load relevant contexts. The saved context can be used as persistent memory across conversations.
How feasible is this idea?
Activity
github-actions commentedon Apr 29, 2026
This issue might be a duplicate of existing issues. Please check:
Consider adding your thoughts to one of those threads if they cover your use case.
github-actions commentedon Jul 3, 2026
To stay organized issues are automatically closed after 60 days of no activity. If the issue is still relevant please open a new one.