rlm
Recursive Language Model for processing large contexts (>50KB). Use for complex analysis tasks where token efficiency matters. Achieves 40% token savings by letting the LLM programmatically explore context via Query() and FINAL() patterns.
What this skill does
# RLM - Recursive Language Model
**RLM** is an inference-time scaling strategy that enables LLMs to handle arbitrarily long contexts by treating prompts as external objects that can be programmatically examined and recursively processed.
- **License:** MIT
- **Repository:** https://github.com/XiaoConstantine/rlm-go
## When to Use
Use `rlm` instead of direct LLM calls when:
- Processing **large contexts** (>50KB of text)
- Token efficiency is important (40% savings on large contexts)
- The task requires **iterative exploration** of data
- Complex analysis that benefits from sub-queries
## Do NOT Use When
- Context is small (<10KB) - overhead not worth it
- Simple single-turn questions
- Tasks that don't require data exploration
## Command Usage
```bash
# Basic usage with context file
~/.local/bin/rlm -context <file> -query "<query>" -verbose
# With inline context
~/.local/bin/rlm -context-string "data" -query "<query>"
# Pipe context from stdin
cat largefile.txt | ~/.local/bin/rlm -query "<query>"
# JSON output for programmatic use
~/.local/bin/rlm -context <file> -query "<query>" -json
```
## Options
| Flag | Description | Default |
|------|-------------|---------|
| `-context` | Path to context file | - |
| `-context-string` | Context string directly | - |
| `-query` | Query to run against context | Required |
| `-model` | LLM model to use | claude-sonnet-4-20250514 |
| `-max-iterations` | Maximum iterations | 30 |
| `-verbose` | Enable verbose output | false |
| `-json` | Output result as JSON | false |
| `-log-dir` | Directory for JSONL logs | - |
## How It Works
RLM uses a Go REPL environment where LLM-generated code can:
1. **Access context** as a string variable
2. **Make recursive sub-LLM calls** via `Query()` for focused analysis
3. **Use standard Go operations** for text processing
4. **Signal completion** with `FINAL()` when done
### The Query() Pattern
```go
// LLM generates code like this inside the REPL:
chunk := context[0:10000]
summary := Query("Summarize the key findings in this text: " + chunk)
// ... iterate through more chunks
FINAL(combinedResult)
```
### The FINAL() Pattern
The LLM signals completion by calling:
- `FINAL("answer")` - Return a string answer
- `FINAL_VAR(variableName)` - Return value of a variable
## Token Efficiency Benefits
For large contexts (>50KB), RLM typically achieves **40% token savings** by:
- Only sending relevant context chunks to sub-queries
- Avoiding repeated full-context processing
- Using programmatic iteration instead of full-context reasoning
## Examples
### Analyze Log Files
```bash
rlm -context server.log -query "Find all unique error patterns and their frequencies"
```
### Process JSON Data
```bash
rlm -context data.json -query "Extract all user IDs with failed transactions" -verbose
```
### Code Analysis
```bash
cat src/*.go | rlm -query "Identify all exported functions and their purposes"
```
## Requirements
- `ANTHROPIC_API_KEY` environment variable must be set
- Binary installed at `~/.local/bin/rlm`
## Installation
```bash
# Quick install
curl -fsSL https://raw.githubusercontent.com/XiaoConstantine/rlm-go/main/install.sh | bash
# Or with Go
go install github.com/XiaoConstantine/rlm-go/cmd/rlm@latest
```
Related in AI Agents
skill-development
IncludedComprehensive meta-skill for creating, managing, validating, auditing, and distributing Claude Code skills and slash commands (unified in v2.1.3+). Provides skill templates, creation workflows, validation patterns, audit checklists, naming conventions, YAML frontmatter guidance, progressive disclosure examples, and best practices lookup. Use when creating new skills, validating existing skills, auditing skill quality, understanding skill architecture, needing skill templates, learning about YAML frontmatter requirements, progressive disclosure patterns, tool restrictions (allowed-tools), skill composition, skill naming conventions, troubleshooting skill activation issues, creating custom slash commands, configuring command frontmatter, using command arguments ($ARGUMENTS, $1, $2), bash execution in commands, file references in commands, command namespacing, plugin commands, MCP slash commands, Skill tool configuration, or deciding between skills vs slash commands. Delegates to docs-management skill for official documentation.
reprompter
IncludedTransform messy prompts into well-structured, effective prompts — single or multi-agent. Use when: "reprompt", "reprompt this", "clean up this prompt", "structure my prompt", rough text needing XML tags and best practices, "reprompter teams", "repromptception", "run with quality", "smart run", "smart agents", multi-agent tasks, audits, parallel work, anything going to agent teams. Don't use when: simple Q&A, pure chat, immediate execution-only tasks. See "Don't Use When" section for details. Outputs: Structured XML/Markdown prompt, quality score (before/after), optional team brief + per-agent sub-prompts, agent team output files. Success criteria: Single mode quality score ≥ 7/10; Repromptception per-agent prompt quality score 8+/10; all required sections present, actionable and specific.
adaptive-compaction
IncludedAdaptive add-on policy and recovery layer that decides WHEN to compact, prune, snapshot, or fork -- replacing fixed-percent auto-compaction across Claude Code, Codex, and MCP-capable hosts. Trigger on auto-compact timing or damage: "when should I compact", "is it safe to compact now or start a fresh session", "auto-compact fires too early/mid-task", "switching to an unrelated task but the window still has space", "context rot", "answers get worse the longer the session runs", "the agent forgot the plan or my decisions after it summarized", "add a layer on top that manages context without changing the agent", raising autoCompactWindow to give the policy room, or installing/tuning a cross-tool compaction policy or PreCompact hook -- even when "compaction" is never said but the problem is context-window pressure or post-summarization memory loss. Do NOT use to summarize a conversation, build RAG, write a summarization prompt (decides WHEN not HOW), or answer max-context-length trivia.
agent-skill-creator
IncludedCreate cross-platform agent skills from workflow descriptions. Activates when users ask to create an agent, automate a repetitive workflow, create a custom skill, or need advanced agent creation. Triggers on phrases like create agent for, automate workflow, create skill for, every day I have to, daily I need to, turn process into agent, need to automate, create a cross-platform skill, validate this skill, export this skill, migrate this skill. Supports single skills, multi-agent suites, transcript processing, template-based creation, interactive configuration, cross-platform export, and spec validation.
llm-wiki
IncludedUse when building or maintaining a persistent personal knowledge base (second brain) in Obsidian where an LLM incrementally ingests sources, updates entity/concept pages, maintains cross-references, and keeps a synthesis current. Triggers include "second brain", "Obsidian wiki", "personal knowledge management", "ingest this paper/article/book", "build a research wiki", "compound knowledge", "Memex", or whenever the user wants knowledge to accumulate across sessions instead of being re-derived by RAG on every query.
skill-master
IncludedAgent Skills authoring, evaluation, and optimization. Create, edit, validate, benchmark, and improve skills following the agentskills.io specification. Use when designing SKILL.md files, structuring skill folders (references, scripts, assets), ingesting external documentation into skills, running trigger evals, benchmarking skill quality, optimizing descriptions, or performing blind A/B comparisons. Keywords: agentskills.io, SKILL.md, skill authoring, eval, benchmark, trigger optimization.