Content scoring & monitoring MCP Server for AI content creators.
Built by @gobuildit — an AI-native content creation community.
- grade_tweet — Grade tweets by engagement + content relevance (S/A/B/C)
- grade_xhs_note — Grade XHS (Xiaohongshu) notes by collect/like ratio + relevance
- grade_content_batch — Batch grade multiple items
- check_relevance — Check if content matches AI/creator topic keywords
- check_domain_exclusion — Auto-filter spam domains (crypto, gambling, MLM, etc.)
- assign_content_tags — Auto-tag content by category (AI/Business/Writing/Growth/Tools)
- get_scoring_config — View current keyword tables and thresholds
- search_x — Search X/Twitter via browser automation
- search_xhs — Search XHS via browser automation
- read_x_timeline — Read X home timeline
- check_cdp_status — Diagnose Chrome CDP connection
Add to your MCP config (~/.claude/mcp_servers.json or Cursor settings):
{
"gobuildit-mcp": {
"type": "stdio",
"command": "uvx",
"args": ["--from", "git+https://github.com/MarkQWu/gobuildit-tools", "gobuildit-mcp"]
}
}That's it. uvx will auto-install and run the server.
See openclaw-skill/gobuildit/ for the skill definition. Copy SKILL.md into your agent's skill directory.
git clone https://github.com/MarkQWu/gobuildit-tools.git
cd gobuildit-tools
# Filter tools only (zero extra deps)
uv pip install -e .
# With monitor tools (adds aiohttp)
uv pip install -e ".[monitor]"
# Run the server
uv run gobuildit-mcpCopy config.example.json to config.json and customize:
{
"relevance_threshold": 3,
"high_confidence_extra": [],
"medium_confidence_extra": [],
"domain_exclude_extra": {},
"cdp_host": "127.0.0.1",
"cdp_port": 9222
}relevance_threshold— minimum score to pass relevance check (default: 3)*_extra— add custom keywords without removing built-in safety filterscdp_host/cdp_port— Chrome DevTools Protocol connection for monitor tools
uv pip install -e ".[dev]"
uv run pytest tests/ -vMIT