Trend Researcher
Trend research and monitoring
Tells you what's genuinely moving in your space and what's just a news cycle. One loud announcement isn't a trend; the same behaviour showing up in three unconnected places is, and it reports the difference. If nothing really changed this week it says so rather than manufacturing three trends.
Add to your OasisConnects with
Connect these once. Oasis holds the connection, and every agent you build after this can reuse it.
ExaRequired
RedditRequired
Hacker NewsRequired
YouTubeRequired
- +Custom toolConnect any other tool from thousands of available integrations.
How it runs
- Foundation
OpenAI Agents
- Model
- GPT-5.6 Sol
- Reasoning
- High effort
- Per run
- Up to 40 turns
- Memory
- Learns across runs
- Browsing
- Can read the web
- Category
- Research
- Trends
- Research
- Scheduled
Frequently asked questions
How does Trend Topic Researcher distinguish a real trend from one loud announcement?
It looks for the same behaviour appearing in three unconnected places; a single announcement is reported as a news cycle rather than promoted to a trend.
What does the trend monitor say when nothing meaningful changed this week?
It says so directly instead of manufacturing three topics to fill a weekly report.
Can Trend Topic Researcher compare signals across Reddit, Hacker News and YouTube?
Those sources are listed toolkits, so the workflow can compare activity across those communities and media channels when assessing movement in a space.
What evidence accompanies a topic that the trend researcher calls genuinely moving?
The report explains the repeated behaviour it found and separates that cross-source pattern from a short-lived news cycle.
Agents like this one
Research Synthesizer
Takes a pile of sources — papers, reports, transcripts, interviews — and produces the synthesis rather than a reading list. It organises by theme, shows where sources agree and whether they're genuinely independent or just citing each other, and never resolves a real conflict by averaging. It ends with the questions the evidence leaves open.
Try this Agent
Oasis PlatformResearchGrowth Researcher
Digs into a growth question until there's an answer with evidence behind it. It makes the question answerable first — which step, which segment, which source — then checks the boring explanations like broken tracking or an ended campaign before reaching for behavioural theories. You get the finding, the numbers supporting it, and an honest confidence level.
Try this AgentOasis PlatformResearchPostHog Product Analyst
Answers product questions from your PostHog data — funnels, retention, feature adoption, cohort comparisons. It verifies the events mean what you think before analysing, because a funnel built on a renamed event produces a confident wrong answer. Every finding comes with sample sizes and the segment breakdown, since a flat overall metric routinely hides one segment collapsing.
Try this AgentOasis PlatformResearchData Analyst
Answers questions from your warehouse or your spreadsheets. It learns the schema and confirms ambiguous definitions before querying, then sanity-checks the result — row counts, reconciliation, fan-out from a bad join — because wrong-but-plausible numbers are the real risk here. You get the answer in one sentence, then the number, then the caveats, with the query shown.
Try this AgentOasis PlatformResearchSearch Performance Reporter
Reports what's actually happening in organic search — which queries and pages gained or lost, and whether a drop is a real ranking change or a tracking artefact. It checks for the technical causes first, because a large unexplained drop is a deindexed page until proven otherwise. Findings are ranked by traffic at stake, with what to do about each.
Try this AgentOasis PlatformResearchCustomer Interview Synthesizer
Reads a batch of recorded interviews and calls and reports the patterns across them. Each theme comes with how many participants raised it, whether they volunteered it or were prompted, and two or three verbatim quotes. It calls out where the findings contradict what the team assumed, which is usually the most valuable part.
Try this AgentOasis PlatformResearchSEO Keyword Research
Finds the queries worth ranking for by weighing volume, difficulty and intent together — a high-volume term that attracts the wrong reader is a cost. It checks what's actually ranking and tells you when the top results all come from domains far stronger than yours. Winners become content briefs with the intent, the questions to answer, and the angle competitors missed.
Try this AgentOasis PlatformSales & MarketingMarket & Industry Brief
A full briefing on a sector you need to speak about credibly tomorrow. It covers who the real players are, how the money actually flows and who makes the buying decision, the regulation where it binds, and what's changing. Market-size figures are dated and sourced, because those numbers get repeated for years past their sell-by date.
Try this AgentOasis PlatformResearchKnowledge Base Q&A
Answers questions from your own documents and cites the source for every claim. It searches broadly first, since the same thing gets described differently across a wiki, a doc and a channel, and shows both when two sources disagree. When the answer genuinely isn't documented it says so and points at who would know, rather than filling the gap with general knowledge.
Try this AgentOasis PlatformResearchMorning Briefing
One message every morning with the day already read for you. It covers what's on the calendar and needs prep, what came in overnight that changes your plans, and which commitments are quietly slipping. It ends with the three things worth doing first, and why.
Try this AgentOasis PlatformProductivity
Put Trend Researcher to work today
Add it to your Oasis, review its setup, and hand it the first task.