intelligence

Build a read on a competitor — what they're actually doing versus what they say, where they're strong, where they're exposed, and what their recent moves imply about their strategy. Use whenever the user is assessing a rival, preparing for a competitive situation, sees a competitor announcement they need to interpret, or asks what a competitor is up to. Trigger on a named competitor plus any strategic, positioning, or threat context.

Segment an unfamiliar market or ecosystem into its structural parts — who the actors are, how value and money move between them, and where the pain concentrates. Use whenever the user is entering a space they don't yet understand, needs a landscape view rather than a company profile, is planning discovery conversations, or asks who the players are in a market. Trigger on "map this space," "landscape," "who's in this market," or any request to understand a market's structure before evaluating anyone in it.

Produce a recurring market intelligence brief — daily or weekly signals filtered for what actually matters to this team, with the "so what" attached. Use whenever the user needs a news or signals digest, is setting up a standing market-monitoring function, has a pile of headlines to filter, or asks what happened in their space this week. Trigger on "news brief," "what's happening in the market," "weekly signals," or any recurring monitoring need.

Size a market with a defensible range rather than a single number — TAM, SAM, and SOM built bottom-up and top-down, reconciled, with every assumption visible and the one that drives the answer named. Use whenever the user asks how big a market is, needs a TAM/SAM/SOM for a plan, deck, or investment case, wants to check a number someone else produced, or is deciding whether an opportunity is large enough to pursue. Trigger on "how big is this market," "TAM," "market size," "is this worth going after," or any request to put a number on an opportunity.

Profile a specific company as a potential partner — what they do, who they serve, how they make money, what they'd want from a partnership, and what would have to be true for it to work. Use whenever the user is evaluating a company for partnership, preparing for a partner conversation, doing build/buy/partner analysis, or asks to look into a specific company. Trigger on a named company plus any partnership, vendor, or collaboration context.

Read across a whole corpus of conversations — meeting notes, interviews, customer calls, briefs — and report what recurs, what conflicts, what is newly emerging, and how the evidence stacks up for and against the team's stated hypotheses, with every claim cited and every theme carrying its independent-source count. Use whenever someone asks what the team is hearing overall, wants to check whether a belief is actually supported, needs a periodic cross-conversation read, or when individual notes have been synthesized but the corpus hasn't. Trigger on "what patterns," "what are we hearing across," "does the evidence support," "themes across conversations," or any request to connect dots across many sources rather than summarize one.

Turn a pile of open questions, graduated signals, and untested assumptions into a prioritized research roadmap — each item framed by the decision it feeds and what "answered" looks like. Use whenever the team has more questions than research capacity, when signals have graduated from a signal log and need a home, when someone asks what we should research next, or when research keeps going to whoever asked loudest. Trigger on "research roadmap," "what should we look into," "research priorities," "research agenda," or any request to organize and rank open questions.

Produce a deep, sourced research brief on a single topic — scope confirmed first, the topic decomposed into lenses so coverage is systematic, every fact sourced and dated, observation kept separate from inference, and the brief ending in what is still unknown. Runs the lenses in parallel through the lens-researcher agent when it is available and sequentially otherwise. Use whenever the user needs to understand a topic, market, practice, or technology deeply enough to make a decision or teach it to others, or asks for "research on," "a brief on," "a deep dive on," "help me understand," or "what do we know about" anything.

Sweep the team's own recent conversations — meeting notes, customer calls, interviews, internal briefs — for research signals, deduplicate them against a running signal log, and flag the ones that have now been seen from more than one independent source. Use whenever a batch of new notes has landed and nobody has asked what they add up to, when setting up a standing "what should we know more about" log, or when the user asks what questions keep coming up. Trigger on "scan the notes," "what are we hearing," "signal log," "what should we be researching," or any request to mine internal notes for open questions rather than action items.