anomalyco/opencode
1
0
Fork 0
mirror of https://github.com/anomalyco/opencode.git synced 2026-07-22 18:25:32 -04:00
Code Issues 5.9k Packages Projects Releases 19 Wiki Activity

[PR #13499] feat: add RLM (Recursive Language Model) engine #14693

New Issue
Open
opened 2026-02-16 18:19:28 -05:00 by yindo · 0 comments
yindo commented 2026-02-16 18:19:28 -05:00
Owner
Copy Link
Copy Source

Original Pull Request: https://github.com/anomalyco/opencode/pull/13499

State: open
Merged: No

This PR adds a JavaScript REPL-based recursive reasoning system that enables LLMs to execute code iteratively in a sandboxed VM environment.

Disclaimer: This code was entirely AI-generated using OpenCode. I haven't had time to review it thoroughly, but I've tested it via nix package install and it works great. Context rebuilding is surprisingly fast.
What's included:

  • Sandboxed vm-based REPL with execution timeout and variable persistence
  • context_query tool for verification during RLM execution
  • Session-level toggle via TUI and rlm_toggle keybind
  • Enabled by default
  • Last summary is always sent along with compacted context, which helps preserve long context and drastically reduces hallucination
    What this enables:
    With RLM, the model can now handle session context questions that were previously impossible — things like:
  • "What was the first error we fixed in this session?"
  • "Which files did we modify 10000 messages ago?"
  • "Summarize all the changes we've made so far using context_query"
  • "What was my original request before we went down this rabbit hole?"
    If you have long sessions where the model starts forgetting earlier context, RLM can help. Let me know how it goes!

ported from: https://github.com/alexzhang13/rlm

**Original Pull Request:** https://github.com/anomalyco/opencode/pull/13499 **State:** open **Merged:** No --- This PR adds a JavaScript REPL-based recursive reasoning system that enables LLMs to execute code iteratively in a sandboxed VM environment. > Disclaimer: This code was entirely AI-generated using OpenCode. I haven't had time to review it thoroughly, but I've tested it via nix package install and it works great. Context rebuilding is surprisingly fast. What's included: - Sandboxed vm-based REPL with execution timeout and variable persistence - context_query tool for verification during RLM execution - Session-level toggle via TUI and rlm_toggle keybind - Enabled by default - Last summary is always sent along with compacted context, which helps preserve long context and drastically reduces hallucination What this enables: With RLM, the model can now handle session context questions that were previously impossible — things like: - "What was the first error we fixed in this session?" - "Which files did we modify 10000 messages ago?" - "Summarize all the changes we've made so far using context_query" - "What was my original request before we went down this rabbit hole?" If you have long sessions where the model starts forgetting earlier context, RLM can help. Let me know how it goes! ported from: https://github.com/alexzhang13/rlm
yindo added the pull-request label 2026-02-16 18:19:28 -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 pull-request
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#14693
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?