New

Agent Research

AI SWOT analysis with live web search & citations

Hand it a company or topic. The agent searches the live web, reads the best sources in full, and writes a SWOT analysis where every claim carries a citation.

Ultra plan · 20 credits per run

How Agent Research works

Every run follows the same five steps. Each one is saved as it completes, so you can watch the research happen in real time.

1 · Plan

The agent breaks your company or topic into research angles and a set of targeted search queries.

2 · Search

It runs live web searches across those queries — current news, filings, and reporting, not stale training data.

3 · Read

It opens the highest-signal sources and reads the full pages, not just the search snippets.

4 · Draft

It writes a structured SWOT and TOWS analysis where every claim is tied to a numbered source.

5 · Review

A final pass checks each claim against its citation and revises anything the sources don't support.

What makes it different

Live web search

Every run pulls current sources at the moment you ask — so the analysis reflects today's reality, not a model's training cutoff.

Reads full pages, not snippets

Instead of skimming a one-line preview, the agent reads the best sources end to end before it writes a single bullet.

Every claim carries a citation

Each strength, weakness, opportunity, and threat is footnoted to the source it came from, so you can verify the reasoning.

See it in action

Browse a SWOT analysis produced by the agent, with citations attached to each point — then run your own on any company or topic.

View an example cited SWOT

Frequently asked questions

What is Agent Research?

Agent Research is an AI agent inside SWOTPal that researches any company or topic on the live web and writes a SWOT analysis where every claim is backed by a cited source. It runs a five-step pipeline — plan, search, read, draft, and review — and saves the result as a normal, editable SWOTPal analysis.

How is it different from a regular AI SWOT generator?

A standard AI SWOT writes from the model's training data and gives you no sources. Agent Research instead searches the live web, reads the best pages in full, and footnotes every claim to a specific source — so the output is current and verifiable rather than a confident guess.

Where do the citations come from?

From the live sources the agent reads during the run. As it searches and opens pages, it tracks which source each fact came from and attaches a numbered citation to every SWOT point. You can open each source from the analysis to check it yourself.

How much does Agent Research cost?

Agent Research is available on the Ultra plan and costs 20 credits per run. Ultra includes 500 credits per month, which is roughly 25 full research runs. Credits are only charged once the analysis is successfully drafted — runs that fail before that point cost nothing.

How long does a run take?

Usually a couple of minutes. Because the agent searches the web, reads multiple full sources, drafts, and then reviews its own work, it takes longer than a one-shot generator — that extra time is what produces the citations.

Is Agent Research available on the free plan?

No. Agent Research is an Ultra-only feature because each run makes live web searches and reads full source pages. Free and Premium plans can still use SWOTPal's standard AI SWOT generator, examples, and templates.

Research any company, fully cited

Let the agent do the digging. You get a structured, source-backed SWOT you can edit, export, and trust.