locoagent  by LocoreMind

AI social media agent automating real browsers

Created 2 weeks ago

New!

514 stars

Top 60.5% on SourcePulse

GitHubView on GitHub
Project Summary

AI-powered social media agent for autonomous operation via real browser automation. It targets users needing sophisticated, human-like interaction, combining LLM-driven decisions with direct browser control via Chrome CDP. The benefit is performing complex tasks like posting, replying, and engaging authentically within actual browser sessions, bypassing bot detection.

How It Works

The system uses an LLM-driven agentic loop interacting with live web pages via agent-browser CLI and Chrome DevTools Protocol (CDP). This ensures operations occur within a real browser instance, using actual login cookies and session fingerprints to evade detection. Key features include a Platform Skill System injecting operation playbooks (e.g., 32+ for X.com) for composite tasks, and a Workflow Engine for deterministic, LLM-free automation pipelines. An Operation Log provides persistent, deduplicated memory of actions. It supports multiple LLM providers via an OpenAI-compatible shim.

Quick Start & Requirements

  • Prerequisites: Bun (Latest), Node.js (>= 18), agent-browser (Latest), Git.
  • Installation: Clone repo, cd locoagent, bun install.
  • Configuration: .env file for LLM provider details (API keys, base URL, model) and SKIP_PERMISSIONS=1 for headless mode.
  • Execution: bun start (interactive), bun start -p "..." (single query, headless), bun start --model ....
  • Browser Setup: Requires Chrome CDP setup (bun run setup-chrome, agent-browser connect 9222).
  • Docs: Workflow development guide at docs/workflow-development-guide.md.

Highlighted Details

  • Real Browser Automation: Uses Chrome CDP with actual login cookies/fingerprints to bypass detection.
  • Platform Skill System: Extensible playbooks (e.g., 32+ for X.com) enable complex, multi-step tasks.
  • Multi-Provider LLM Support: Integrates with OpenAI-compatible APIs (OpenRouter, DeepSeek, Ollama, Anthropic, etc.).
  • Persistent Operation Log: Deduplicates actions across sessions.
  • Workflow Engine: Manages deterministic, scripted automation pipelines.

Maintenance & Community

The README does not specify notable contributors, sponsorships, or community channels (e.g., Discord, Slack).

Licensing & Compatibility

MIT License, generally permitting commercial use and modification.

Limitations & Caveats

Setup involves Chrome CDP management and .env configuration. Reliance on real browser automation makes it susceptible to platform UI changes. The Bun runtime dependency may be a consideration.

Health Check
Last Commit

1 week ago

Responsiveness

Inactive

Pull Requests (30d)
1
Issues (30d)
0
Star History
519 stars in the last 14 days

Explore Similar Projects

Feedback? Help us improve.