aiwg
Version:
Cognitive architecture for AI-augmented software development with structured memory, ensemble validation, and closed-loop correction. FAIR-aligned artifacts, 84% cost reduction via human-in-the-loop, standards adopted by 100+ organizations.
76 lines (75 loc) • 1.82 kB
JSON
{
"id": "rlm",
"type": "addon",
"name": "Recursive Language Models",
"version": "1.0.0",
"description": "RLM-inspired recursive context decomposition for processing arbitrarily large codebases and document corpora through programmatic sub-agent delegation",
"core": false,
"autoInstall": false,
"author": "AIWG Contributors",
"license": "MIT",
"repository": "https://github.com/jmagly/aiwg",
"keywords": [
"rlm",
"recursive",
"long-context",
"decomposition",
"sub-agent",
"context-management",
"chunking",
"fan-out",
"batch"
],
"researchFoundation": {
"REF-089": "Recursive Language Models (Zhang et al., 2026) — REPL-based recursive decomposition for 10M+ token processing",
"REF-018": "ReAct — TAO loop structure underlying RLM's iterative REPL",
"REF-022": "AutoGen — Sub-agent delegation patterns parallel to llm_query()",
"REF-015": "Self-Refine — Iterative refinement with best-output selection"
},
"entry": {
"agents": "agents/",
"commands": "commands/",
"skills": "skills/",
"rules": "rules/",
"schemas": "schemas/",
"templates": "templates/",
"docs": "docs/"
},
"agents": [
"rlm-agent"
],
"commands": [
"rlm-query",
"rlm-batch",
"rlm-status"
],
"skills": [
"rlm-mode"
],
"rules": [
"rlm-context-management"
],
"schemas": [
"rlm-task-tree",
"rlm-state",
"rlm-trajectory",
"rlm-cost"
],
"templates": [
"cost-report"
],
"dependencies": {
"required": [],
"optional": ["ralph", "aiwg-utils"]
},
"configuration": {
"defaults": {
"maxDepth": 3,
"maxSubCalls": 20,
"defaultSubModel": "sonnet",
"budgetTokens": 500000,
"parallelSubCalls": true,
"persistState": true
}
}
}