{
  "id": 191710,
  "title": "Meta launches Muse Code, an AI agent for large code bases",
  "url": "https://urgent.news/2026/08/05/meta-launches-muse-code-an-ai-agent-for-large-code-bases",
  "topic": "ai",
  "section": "AI",
  "published": "2026-08-05T21:21:28.000Z",
  "source": {
    "name": "TechCrunch",
    "slug": "techcrunch",
    "url": "https://techcrunch.com/2026/08/05/meta-launches-muse-code-an-ai-agent-for-large-code-bases/"
  },
  "original_language": "en",
  "account": "Meta, previously seen as a latecomer in the AI landscape, is making efforts to level the playing field. The tech giant recently unveiled a new tool called Muse Code, an Artificial Intelligence agent designed for software developers seeking help with intricate tasks in expansive code repositories. Muse Code, currently in its beta phase, is capable of completing extensive software engineering assignments across vast software repositories, according to Meta CEO Mark Zuckerberg's announcement on social media.\n\nThese tasks encompass planning modifications, composing code, and confirming the outcomes. Muse Code, which can be installed using a single command, is driven by Meta's earlier released coding model, Muse Spark. It tackles large projects by deploying its own agents that operate concurrently. Zuckerberg explained that when a job is substantial enough, it dispatches separate sub-agents to work independently in distinct worktrees. Importantly, the programmer's current working copy remains untouched. In testing, Meta demonstrated Muse Code simultaneously constructing six features for a game without any conflicts.\n\nThis development aims to enhance Meta's competitiveness in the AI sector and make it more cost-effective compared to rivals like OpenAI and its coding agent Codex, and Anthropic with Claude Code. Alexandr Wang, Meta's AI chief overseeing Meta Superintelligence Labs, expressed optimism about Muse Code, stating that it could be a highly beneficial option for numerous workflows and use cases, particularly from a cost standpoint. Meta has been aggressively expanding its AI footprint by investing heavily in its development. In June, the company broadened its AI scope beyond its primary focus on AI-driven advertising and entered the enterprise AI market with an agent designed for customer service and support.",
  "summary": "Meta expanded its AI coding offerings with a new agent that, it promises, can handle complex tasks with complex software.",
  "key_points": [],
  "editors_take": null,
  "illustration": null,
  "coverage": {
    "outlets": 1,
    "also_reported_by": []
  },
  "ai_generated": true,
  "disclaimer": "Summaries, key points and the editor’s take are written by software from other outlets’ reporting and may contain errors — always check the linked original."
}