[{"data":1,"prerenderedAt":4},["ShallowReactive",2],{"readme:autoresearch":3},"\u003Cdiv align=\"center\">\u003Ch1>Autoresearch\u003C\u002Fh1>\n\u003Cp>\u003Cstrong>Turn \u003Ca href=\"https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code\" rel=\"nofollow ugc noopener\">Claude Code\u003C\u002Fa>, \u003Ca href=\"https:\u002F\u002Fopencode.ai\" rel=\"nofollow ugc noopener\">OpenCode\u003C\u002Fa>, or \u003Ca href=\"https:\u002F\u002Fdevelopers.openai.com\u002Fcodex\" rel=\"nofollow ugc noopener\">OpenAI Codex\u003C\u002Fa> into a relentless improvement engine.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cp>Based on \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fkarpathy\u002Fautoresearch\" rel=\"nofollow ugc noopener\">Karpathy's autoresearch\u003C\u002Fa> — constraint + mechanical metric + autonomous iteration = compounding gains.\u003C\u002Fp>\n\u003Cp>\u003Ca href=\"https:\u002F\u002Fdocs.anthropic.com\u002Fen\u002Fdocs\u002Fclaude-code\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FClaude_Code-Skill-blue?logo=anthropic&logoColor=white\" alt=\"Claude Code Skill\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fopencode.ai\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FOpenCode-Skill-purple\" alt=\"OpenCode\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fdevelopers.openai.com\u002Fcodex\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FCodex-Skill-green?logo=openai&logoColor=white\" alt=\"Codex\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuditgoenka\u002Fautoresearch\u002Freleases\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002Fversion-2.2.1-blue.svg\" alt=\"Version\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuditgoenka\u002Fautoresearch\u002Fblob\u002FHEAD\u002FLICENSE\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FLicense-MIT-green.svg\" alt=\"License: MIT\" \u002F>\u003C\u002Fa>\u003C\u002Fp>\n\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fkarpathy\u002Fautoresearch\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FBased_on-Karpathy's_Autoresearch-orange\" alt=\"Based on\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fx.com\u002Fintent\u002Ffollow?screen_name=iuditg\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FFollow-@iuditg-000000?style=flat&logo=x&logoColor=white\" alt=\"Follow @iuditg\" \u002F>\u003C\u002Fa>\n\u003Ca href=\"https:\u002F\u002Fpaypal.me\u002Fuditgoenka\" rel=\"nofollow ugc noopener\">\u003Cimg src=\"https:\u002F\u002Fimg.shields.io\u002Fbadge\u002FSupport-PayPal-00457C?style=flat&logo=paypal&logoColor=white\" alt=\"Support\" \u002F>\u003C\u002Fa>\u003C\u002Fp>\n\u003Cbr \u002F>\u003Cp>\u003Cem>\"Set the GOAL → The agent runs the LOOP → You wake up to results\"\u003C\u002Fem>\u003C\u002Fp>\n\u003Cp>\u003Cem>You don't need AGI. You need a goal, a metric, and a loop that never quits.\u003C\u002Fem>\u003C\u002Fp>\n\u003Cp>\u003Cstrong>Supports Claude Code, OpenCode, and OpenAI Codex. 14 commands. 9 safety hooks. 95% fewer tokens per invocation.\u003C\u002Fstrong>\u003C\u002Fp>\n\u003Cblockquote>\n\u003Cp>\u003Cstrong>v2.2.0 — Autonomous Orchestrator:\u003C\u002Fstrong> Type a plain-language goal to \u003Ccode>\u002Fautoresearch\u003C\u002Fcode> and it classifies your goal, derives a Success predicate, confirms it once, then loops across subcommands until done. No manual chaining required. \u003Ccode>Metric:\u003C\u002Fcode>\u002F\u003Ccode>Verify:\u003C\u002Fcode> invocations run the classic loop unchanged. See \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuditgoenka\u002Fautoresearch\u002Fblob\u002FHEAD\u002Fguide\u002Fautoresearch-orchestrator.md\" rel=\"nofollow ugc noopener\">guide\u002Fautoresearch-orchestrator.md\u003C\u002Fa>.\u003C\u002Fp>\n\u003C\u002Fblockquote>\n\u003Cbr \u002F>\u003Cp>\u003Ca href=\"#how-it-works\" rel=\"nofollow ugc noopener\">How It Works\u003C\u002Fa> · \u003Ca href=\"#commands\" rel=\"nofollow ugc noopener\">Commands\u003C\u002Fa> · \u003Ca href=\"#quick-start\" rel=\"nofollow ugc noopener\">Quick Start\u003C\u002Fa> · \u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fuditgoenka\u002Fautoresearch\u002Fblob\u002FHEAD\u002Fguide\u002F\" rel=\"nofollow ugc noopener\">Guides\u003C\u002Fa> · \u003Ca href=\"#faq\" rel=\"nofollow ugc noopener\">FAQ\u003C\u002Fa>\u003C\u002Fp>\n\u003C\u002Fdiv>\u003Chr \u002F>\n\u003Cpre>\u003Ccode>     PLAN             LOOP            DEBUG             FIX             SECURE            SHIP\n ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐\n │   Goal   │     │  Modify  │     │   Find   │     │   Fix    │     │  STRIDE  │     │  Stage   │\n │  Metric  │────▶│  Verify  │────▶│   Bugs   │────▶│  Errors  │────▶│  OWASP   │────▶│  Deploy  │\n │  Scope   │     │Keep\u002FDrop │     │  Trace   │     │  Repair  │     │ Red Team │     │ Release  │\n └──────────┘     └──────────┘     └──────────┘     └──────────┘     └──────────┘     └──────────┘\n \u002Fautoresearch:   \u002Fautoresearch    \u002Fautoresearch:   \u002Fautoresearch:   \u002Fautoresearch:   \u002Fautoresearch:\n   plan                              debug            fix              security         ship\n\n ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐\n │  Probe   │     │ Scenario │     │ Predict  │     │  Reason  │\n │ Require- │     │   Edge   │     │ 5-Expert │     │  Debate  │\n │  ments   │     │  Cases   │     │  Swarm   │     │ Converge │\n └──────────┘     └──────────┘     └──────────┘     └──────────┘\n \u002Fautoresearch:   \u002Fautoresearch:   \u002Fautoresearch:   \u002Fautoresearch:\n   probe            scenario         predict          reason\n\n ┌──────────┐     ┌──────────┐     ┌──────────┐     ┌──────────┐\n │  Learn   │     │ Improve  │     │   Eval   │     │ Baseline │\n │   Docs   │     │ Research │     │ Analyze  │     │   Diff   │\n │   Gen    │     │   PRDs   │     │ Results  │     │ Verdict  │\n └──────────┘     └──────────┘     └──────────┘     └──────────┘\n \u002Fautoresearch:   \u002Fautoresearch:   \u002Fautoresearch:   \u002Fautoresearch:\n   learn            improve          evals            regression\n\u003C\u002Fcode>\u003C\u002Fpre>\n\u003Chr \u002F>\n\u003Ch2>Why This Exists\u003C\u002Fh2>\n\u003Cp>\u003Ca href=\"https:\u002F\u002Fgithub.com\u002Fkarpathy\u002Fautoresearch\" rel=\"nofollow ugc noopener\">Karpathy's autoresearch\u003C\u002Fa> demonstrated that a 630-line Python script cou\u003C\u002Fp>\n",1784240406343]