GUIDES
Guides to finding, checking and using trends
Developer recipes for the V3 API and MCP server, and guides to deciding what to publish from the evidence. Each is dated and says where its evidence stops.
ANSWER FIRST
The short answer
The developer recipes show the V3 API and MCP server at work: connecting an agent, calling it from Node.js or Python, comparing history, judging evidence and budgeting credits. The other guides cover planning and agent workflows: what to post today, how to start on Reddit and X, what to make when there is no clear trend, and how to run Claude as a head of growth.
Developer recipes for the V3 API
- Use a trends API from Node.js. Query saved global observations, inspect evidence and rank a niche with explicit credit limits in a small server-side Node.js integration.
- Ask a trend question with Python. Build a small Python consumer that keeps an answer, exact source references, coverage and credits together without silently fetching new research.
- Connect an agent to TrendsFast through MCP. Understand the remote MCP endpoint, scoped OAuth or API keys, bounded query tools and free saved-result reads before connecting an agent.
- Measure trend growth from historical observations. Compare the same source entity and metric across retained timestamps, keep corrections visible and explain missing baselines without inventing a trend line.
- Define a niche without losing useful adjacent trends. Use confirmed audience, jobs, problems and exclusions to rank shared observations across CORE, ADJACENT and GLOBAL research rings.
- Audit trend evidence before your agent uses it. Separate source references, measured counters, permitted excerpts, media inspection and AI-search citations so an agent can explain what it knows.
- Budget trend API credits without duplicate queries. Use quotes, finite maximums, durable request IDs and free result recovery to control cached-query spending across REST and MCP.
Planning, evidence and agent workflows
- Start on Reddit and X with no audience. Choose a narrow audience, answer a real question and share something useful. A worked founder example, posting checks and a simple way to learn from results.
- What is social and search trend intelligence?. A practical definition of social and search trend intelligence, its evidence boundary, and how an agent turns it into one useful piece.
- What should I post today for my product?. A five-step decision framework for choosing what to post today without forcing a trend or losing the evidence behind the idea.
- Turn Claude into your Head of Growth with TrendsFast. Connect TrendsFast to Claude, paste one set of instructions and ask one question every Monday: what to post this week, and where, from what your customers are stuck on and inside each community's rules.
- Add trend research to your SaaS without building five integrations. What it takes to put live trend research inside your own product or agent: the one call that answers across search, AI search, conversations and video, what it costs, and what it refuses to claim.
- Trend intelligence for AI agents. How an AI agent reads your brand context, makes one capped niche query and rereads the evidence for free through MCP and REST.
- What to create when there is no clear trend. Why no clear trend is an honest result, how the evidence floor works, and when an evergreen insight can still support useful content.
Elsewhere
The docs hold the API reference and quickstart, the comparisons weigh TrendsFast against other trend tools, and the blog explains how the evidence is judged.