anomalyco/opencode
1
0
Fork 0
mirror of https://github.com/anomalyco/opencode.git synced 2026-07-22 10:15:31 -04:00
Code Issues 5.9k Packages Projects Releases 19 Wiki Activity

[FEATURE]:optional lightweight RAG-based context extension (LLM RAG Booster integration) #9043

New Issue
Open
opened 2026-02-16 18:11:30 -05:00 by yindo · 0 comments
yindo commented 2026-02-16 18:11:30 -05:00
Owner
Copy Link
Copy Source

Originally created by @Tryboy869 on GitHub (Feb 11, 2026).

Originally assigned to: @thdxr on GitHub.

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

📝 Description

Problem

OpenCode works very well for interactive coding, but it is currently limited by the native context window of the underlying LLM when working with:
large codebases
multi-repo projects
long-lived sessions
extensive documentation or specs
Even with good chunking, relevant context is often dropped once the token limit is reached.

Proposed solution

I would like to propose an optional integration of a lightweight, dependency-free RAG-style context extension called LLM RAG Booster:

https://github.com/Tryboy869/llm-rag-booster-allpath�
Key characteristics:
Pure Python, MIT licensed
No vector database
No embeddings
In-memory compression + keyword indexing
Works with any LLM provider (OpenAI, Anthropic, Groq, Ollama, etc.)
Designed to extend usable context ~15–60× beyond native limits

Can be used as a standalone pre-processing / retrieval layer
The idea would not be to replace existing OpenCode behavior, but to optionally allow:
loading large codebases or docs into a compressed memory layer
retrieving top-k relevant chunks at inference time
injecting that context into the existing OpenCode agent flow
Why this fits OpenCode

Provider-agnostic (matches OpenCode philosophy)
Client/server friendly (can run as a sidecar process)
No heavy infra (no DB, no services)
Optional and opt-in

Useful for large monorepos and long sessions
Complements LSP + agent-based workflows
Possible integration approach (high level)
External helper process or plugin
Feed retrieved chunks into the agent prompt before inference

Toggleable via config (e.g. use_extended_context: true)
No UI changes required initially
I’m happy to adapt the project to better fit OpenCode’s architecture if there is interest.

Questions

Would this be considered in-scope as an optional feature or plugin?
Is there an existing extension point better suited for this?
Would maintainers prefer this as a standalone plugin rather than core integration?

Originally created by @Tryboy869 on GitHub (Feb 11, 2026). Originally assigned to: @thdxr on GitHub. ### Feature hasn't been suggested before. - [x] I have verified this feature I'm about to request hasn't been suggested before. ### Describe the enhancement you want to request 📝 Description **_Problem_** OpenCode works very well for interactive coding, but it is currently limited by the native context window of the underlying LLM when working with: large codebases multi-repo projects long-lived sessions extensive documentation or specs Even with good chunking, relevant context is often dropped once the token limit is reached. **Proposed solution** I would like to propose an optional integration of a lightweight, dependency-free RAG-style context extension called LLM RAG Booster: https://github.com/Tryboy869/llm-rag-booster-allpath� Key characteristics: Pure Python, MIT licensed No vector database No embeddings In-memory compression + keyword indexing Works with any LLM provider (OpenAI, Anthropic, Groq, Ollama, etc.) Designed to extend usable context ~15–60× beyond native limits Can be used as a standalone pre-processing / retrieval layer The idea would not be to replace existing OpenCode behavior, but to optionally allow: loading large codebases or docs into a compressed memory layer retrieving top-k relevant chunks at inference time injecting that context into the existing OpenCode agent flow Why this fits OpenCode Provider-agnostic (matches OpenCode philosophy) Client/server friendly (can run as a sidecar process) No heavy infra (no DB, no services) Optional and opt-in Useful for large monorepos and long sessions Complements LSP + agent-based workflows Possible integration approach (high level) External helper process or plugin Feed retrieved chunks into the agent prompt before inference Toggleable via config (e.g. use_extended_context: true) No UI changes required initially I’m happy to adapt the project to better fit OpenCode’s architecture if there is interest. **Questions** Would this be considered in-scope as an optional feature or plugin? Is there an existing extension point better suited for this? Would maintainers prefer this as a standalone plugin rather than core integration?
yindo added the discussion label 2026-02-16 18:11:30 -05:00
Sign in to join this conversation.
No Branch/Tag Specified
Branches Tags
Labels
Clear labels
acp
bad
bug
discussion
docs
documentation
enhancement
extension
good first issue
growth
help-wanted
mcp
model-problem
needs:compliance
nix
opentui
perf
pull-request
Mirrored from GitHub Pull Request
 tui
web
windows
zen
No Label discussion
Milestone
No items
No Milestone
Projects
Clear projects
No project
Assignees
Clear assignees
yindo
No Assignees
1 Participants
Notifications
Due Date
No due date set.
Dependencies

No dependencies set.

Reference: anomalyco/opencode#9043
Delete Branch "%!s()"

Deleting a branch is permanent. Although the deleted branch may continue to exist for a short time before it actually gets removed, it CANNOT be undone in most cases. Continue?