Skills DirectorySkills Directory
SkillsLearnSecurityCategoriesDocsCommunityBlog
Sign InSubmit Skill
Skills Directory

Security-tested agent skills for Claude, coding agents, and AI workflows.

Directory

  • Browse Skills
  • All Skills A–Z
  • Claude Skills
  • Claude Code Skills
  • Agent Skills
  • Categories
  • Submit a Skill

Learn

  • Learn Hub
  • Install Claude Skills
  • Write SKILL.md
  • Skills vs MCP
  • Directories Compared

Security

  • Security
  • Methodology
  • Secure Claude Skills
  • Security Badges

Company

  • About
  • Community
  • Blog
  • API Docs
  • Advertise

2026 Skills Directory. All rights reserved.

Back to skills

Rlm Recursive Agent

ASecurity

Use when implementing Prime Agent / Recursive Language Model (RLM) patterns — self-improving coding agent that beats human baseline on ARC-AGI3 (95.5% MIT), runs in terminal like Claude Code/Codex but with recursive sub-agents and Pluggable LLM backends (Claude, Codex, GLM). Triggers on "RLM", "Prime Agent", "recursive language model", "ARC-AGI", "self-improving agent", "agente recursivo".

2 stars
0 votes
0 copies
0 views
Added 9/19/2026
ai-agentsrustgobashgitapibackend

Works with

claude codeterminalapi

Security Analysis

A96/100
mediumInstalls packages at runtime which could introduce malicious dependencies

Scanned 9/19/2026

Install to Claude Code

$npx -y skills add majinmagros/magros.ai-skills --skill rlm-recursive-agent --agent claude-code

Installs into .claude/skills of the current project.

Are you the author of Rlm Recursive Agent?

Add the live security badge to your README — it updates automatically with every re-scan.

Security grade badge for Rlm Recursive Agent
[![Security: A — Skills Directory](https://www.skillsdirectory.com/api/skills/majinmagros-rlm-recursive-agent/badge)](https://www.skillsdirectory.com/skills/majinmagros-rlm-recursive-agent)

More formats (shields.io, HTML) on the badges page.

Download Zip
Files
SKILL.md
---
name: rlm-recursive-agent
description: Use when implementing Prime Agent / Recursive Language Model (RLM) patterns — self-improving coding agent that beats human baseline on ARC-AGI3 (95.5% MIT), runs in terminal like Claude Code/Codex but with recursive sub-agents and Pluggable LLM backends (Claude, Codex, GLM). Triggers on "RLM", "Prime Agent", "recursive language model", "ARC-AGI", "self-improving agent", "agente recursivo".
metadata:
  origin: ECC
---

# RLM Recursive Agent — Prime Agent

Based on **Prime Intellect Prime Agent** (video `P6X037tssiE` 2026-08-13, ai-code-king): open-source MIT terminal agent built on **RLM (Recursive Language Model)** architecture that scores **95.5% on ARC-AGI 3** (human expert baseline 95.4%).

> Fonte: transcrição `P6X037tssiE.en.dedup.txt` (ai-code-king/RELATORIO.md 2026-08-26) — validado contra repo `prime-intellect/prime-agent` (MIT).

## Quando usar

- Agente que se auto-melhora entre runs (self-improvement) sem mudar modelo base
- ARC-AGI3 / tarefas de raciocínio recursivo onde decomposição plana falha
- Terminal harness que precisa plugar qualquer LLM (Claude, Codex, GLM-4.5, Kimi) via API key ou subscription existente
- Alternativa open-source a Claude Code / Codex para times que querem harness auditável

## Quando NÃO usar

- Tarefa single-file trivial (use `claude -p` direto)
- Já está em `agentic-os` / `gan-style-harness` e não precisa de loop recursivo
- Precisa de execução 100% local sem API (RLM ainda chama LLM externo)

## Arquitetura RLM (vs harness plano)

```
Harness plano:  prompt → LLM → tools → resposta
RLM:            prompt → LLM_root
                          ├─→ sub-problema_1 → LLM_leaf → tools → synthesis
                          ├─→ sub-problema_2 → LLM_leaf → ...
                          └─→ sub-problema_N → ...
                          ↺  root re-avalia síntese, gera nova decomposição (loop até convergência)
```

- **Root** decompõe e orquestra; **leaves** executam sub-tarefas com contexto isolado
- **Síntese recursiva**: folhas retornam solução parcial, root integra e decide se precisa recursar
- **Self-improvement**: traces de falha/sucesso viram prompt para próxima iteração (sem fine-tuning)

## Uso

### Instalação (MIT)

```bash
git clone https://github.com/prime-intellect/prime-agent
cd prime-agent && npm i  # ou cargo install se Rust
export ANTHROPIC_API_KEY=... # ou OPENAI_API_KEY / ZAI_API_KEY para GLM
prime-agent --model claude-sonnet-4 --task "refatore o auth para JWT"
```

### Com modelos plugáveis

```bash
# Via subscription existente (Claude/Codex)
prime-agent --provider claude --model claude-sonnet-4
prime-agent --provider openai --model gpt-4o

# Via API barata (GLM/Kimi)
prime-agent --provider zai --model glm-4.5 --api-key $ZAI_API_KEY
prime-agent --provider moonshot --model kimi-k2
```

### No harness ECC

Attribution

majinmagrosmajinmagros
View sourceMore from majinmagros →
SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Is this your skill, or is something wrong with this listing? Request removal or report an issue. Author removals are honored within 72 hours.

Comments (0)

No comments yet. Be the first to comment!

SSkills DirectorySkills Directory

Ship a skill? Prove it's safe.

Free 120-pattern security scan, letter grade, and an embeddable README badge.

Submit a skill

Related Skills

Caveman

Ultra-compressed communication mode. Cuts token usage ~75% by speaking like caveman while keeping full technical accuracy. Supports intensity levels: lite, full (default), ultra, wenyan-lite, wenyan-full, wenyan-ultra. Use when user says "caveman mode", "talk like caveman", "use caveman", "less tokens", "be brief", or invokes /caveman. Also auto-triggers when token efficiency is requested.

1023331 votes

Hyperplan

Adversarial multi-agent planning skill. Self-orchestrates 5 hostile category members (unspecified-low, unspecified-high, deep, ultrabrain, artistry) via team-mode for ruthless cross-critique debate, distills only the defensible insights, then MANDATORILY hands the distilled insight bundle to the `plan` agent for executable plan formalization. Use when planning needs maximum rigor and surfacing of weak assumptions, blind spots, and over-engineering. Triggers: 'hyperplan', 'hpp', '/hyperplan', ...

686011 votes

Mcp Code Execution

Routes multi-tool workflows through MCP servers for large datasets and pipelines. Use when Bash tool overhead is limiting throughput on data-heavy tasks.

3331 votes

catchup

Recovers prior coding-agent session context by running `catchup <agent> --since-compact`, which extracts a clean summary of a previous Codex, Claude Code, Antigravity, OpenCode, or Pi Agent session. Use when the user says "catch up", "what did the last session do", "get me up to speed", "I switched agents", or asks to recover/summarize a previous session before continuing. Do NOT use for the current conversation, git history, or any non-agent log.

611 votes

math-skill

A comprehensive mathematical reasoning skill for AI assistants — handles arithmetic to research-level problems with rigorous step-by-step reasoning, systematic verification, and transparent uncertainty handling

381 votes
View all in ai-agents →