Two browser automation solutions purpose-built for AI agents, both born in January 2026, taking radically different architectural approaches.

A quantitative and qualitative comparison of Fulcrum, Vibe Kanban, Auto-Claude, and OpenClaw

The AI-assisted development tool space is exploding. In the span of a few months, multiple projects have emerged with overlapping but distinct visions for how developers should interact with AI coding agents. I took four of them — Fulcrum, Vibe Kanban, Auto-Claude, and OpenClaw — and put them under the microscope using scc for code metrics, git log for history, and manual code review for everything else.

What follows is a framework-driven comparison across eight dimensions, from raw LOC counts to architectural philosophy to "are they punching above their weight?"

• SSH to citadel (5.78.40.219) hung — TCP connected but no SSH banner returned
• Hetzner VNC console showed continuous OOM killer messages:

  Memory cgroup out of memory: Killed process XXXX (windmill)
  Memory cgroup out of memory: Killed process XXXX (uv)
  

• CPU pinned at ~1000-1500% (OOM kill/respawn loop)

You help users explore and analyze data from Augment Risk's CRM datalake. Use the queryDucklake action to run SQL queries (DuckDB dialect).

Be proactive, not inquisitive. Infer user intent from context and act on it. Do not bombard users with clarifying questions. Make reasonable assumptions and execute. If you're wrong, the user will guide or correct you. It's better to attempt something useful than to ask for permission.

Always check schema before querying. Before running any analytics query, first run DESCRIBE main.table_name for every table you plan to query or join. Column names and types vary - never assume. This prevents errors and ensures accurate results.

You're paying for ChatGPT Plus/Pro ($20-200/month) AND buying OpenAI API tokens separately for AI-assisted coding. That's double-dipping.

opencode-openai-codex-auth lets you authenticate OpenCode with your existing ChatGPT subscription, giving you access to GPT-5.x models for coding tasks without purchasing additional API credits.

Bottom line: The Claude Max plan delivers $800-1,600+ worth of API tokens - and you can use it programmatically via the Agent SDK.

Date: 2025-12-31 Budget Used: ~$5-10 (estimated from 2x ~20min A10G training runs + eval)

MCP-DSL demonstrates significant advantages over JSON-RPC for small model fine-tuning:

| Metric | MCP-DSL | JSON-RPC | Advantage |

A simple, robust way to maintain persistent SSH tunnels that automatically reconnect after network interruptions, sleep/wake cycles, or server restarts.

Instead of manually running SSH tunnels or using third-party tools like autossh, we use macOS's built-in launchd service manager. It handles:

• Automatic startup on login
• Automatic restart when the connection drops

This guide hardens Docker container isolation on a system running Dokploy with untrusted public templates.