name	technical-trends-discovery
description	Discover what's trending in AI/tech by scanning thought leaders on X, then deep-researching the #1 trend with structured output (docs URLs, package versions, key concepts). Use when you want to stay current on AI engineering trends without manual research.

Technical Trends Discovery

X/Twitter is where thought leaders share real-time opinions and insights. The xAI API excels at searching and analyzing X posts to identify what's actually important right now. Once we identify THE trend, Tavily Research extracts structured metadata including latest package versions, key concepts, documentation URLs, and insights.

Two-step automated pipeline for discovering and deeply researching the most important AI/tech trend:

X Search (xAI API) → Find what thought leaders are discussing, identify #1 trend
Deep Research (Tavily API) → Comprehensive research on the topic

Quick Start

# Run full pipeline: X → Research → Structured JSON
python .claude/skills/technical-trends-discovery/scripts/discover_trends.py

# X discovery only (skip Tavily research)
python .claude/skills/technical-trends-discovery/scripts/discover_trends.py --x-only

# Custom handles and date range
python .claude/skills/technical-trends-discovery/scripts/discover_trends.py \
  --handles karpathy simonw swyx \
  --days 7

Output Format

Results are saved to trends-reports/ at the repo root as a single consolidated JSON file:

trends-reports/
└── trends_2025-01-06_143022/
    └── report.json         # All results in one file

report.json Structure

{
  "meta": {
    "generated_at": "2025-01-06T14:30:22.123456",
    "pipeline": "x_discovery → tavily_research",
    "sources_count": 15
  },
  "x_discovery": {
    "content": "# X Trends Analysis\n\nThe #1 trend identified...",
    "citations": ["https://x.com/..."]
  },
  "research": {
    "trend": {
      "name": "Model Context Protocol",
      "summary": "An open standard for sharing context between AI tools...",
      "why_important": "MCP is emerging as the standard...",
      "docs_url": "https://modelcontextprotocol.io",
      "github_repo": "https://github.com/modelcontextprotocol",
      "quickstart": {
        "prerequisites": ["Python 3.10+"],
        "install_commands": "pip install mcp==1.25.0",
        "hello_world_code": "from mcp import Client...",
        "expected_output": "Connected to MCP server"
      },
      "use_cases": [...],
      "common_pitfalls": [...],
      "key_packages": [
        {"name": "mcp", "latest_version": "1.25.0", "package_manager": "pip"}
      ],
      "key_concepts": ["Resources", "Tools", "Prompts"],
      "additional_resources": [...]
    },
    "meta": {"research_date": "2025-01-06"}
  },
  "sources": [
    {"url": "https://...", "title": "MCP Documentation"}
  ]
}

Default Thought Leaders

Handle	Person
hwchase17	Harrison Chase (LangChain)
rlancemartin	Lance Martin (LangChain)
simonw	Simon Willison
karpathy	Andrej Karpathy
cherny	Boris Cherny
swyx	Swyx
alexalbert__	Alex Albert (Anthropic)

CLI Options

Option	Default	Description
`--handles`, `-H`	7 AI leaders	X handles to search
`--days`, `-d`	20	Days back to search

Environment Variables

export XAI_API_KEY="your-xai-key"      # Required for X search
export TAVILY_API_KEY="your-tavily-key" # Required for research

Python Usage

from discover_trends import discover_trends

# Run full pipeline
results = discover_trends(
    handles=["karpathy", "simonw", "swyx"],
    days_back=14,
)

# Access results
print(results["x_trends"]["content"])       # X discovery markdown
print(results["research"]["content"])       # Structured JSON with trend data
print(results["output_dir"])                # Where files were saved

technical-trends-discovery

Install Skill

SKILL.md