rlm-cli 1.3.0

Recursive Language Model (RLM) REPL for Claude Code - handles long-context tasks via chunking and recursive sub-LLM calls
Documentation 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34

---
title: rlm-rs
description: Recursive Language Model (RLM) REPL for Claude Code - handles long-context tasks via chunking and recursive sub-LLM calls.
template: splash
hero:
  title: rlm-rs
  tagline: Recursive Language Model REPL for Claude Code — long-context tasks via chunking and recursive sub-LLM calls
  actions:
    - text: Get Started
      link: /rlm-rs/getting-started/features/
      icon: right-arrow
      variant: primary
    - text: View on GitHub
      link: https://github.com/zircote/rlm-rs
      icon: external
      variant: minimal
---

import { Card, CardGrid } from '@astrojs/starlight/components';

<CardGrid>
  <Card title="Recursive Chunking" icon="rocket">
    Break large files into manageable chunks and process them recursively with sub-LLM calls for comprehensive analysis.
  </Card>
  <Card title="Hybrid Search" icon="magnifier">
    Combines BM25 text search with vector similarity using Reciprocal Rank Fusion for precise context retrieval.
  </Card>
  <Card title="Embedded Embeddings" icon="puzzle">
    Ships with BGE-M3 model via fastembed — no external API calls needed for vector operations.
  </Card>
  <Card title="Claude Code Plugin" icon="setting">
    Integrates as a Claude Code skill for seamless use within your existing AI-assisted development workflow.
  </Card>
</CardGrid>