Stocksbrew Review: Stock Research Your AI Agent Can Actually Read

Most stock research sites are built for a human with a browser. Stocksbrew is built for a human with a browser and an AI agent, and the second half is the interesting part. It exposes its own research through a Model Context Protocol interface, so Claude, Codex, or Cursor can read the same fundamentals and watchlist you see on the page. For a research step that happens before a trade, that changes the workflow more than another dashboard would.

A note on scope first. ChartRecap is a journal, not a signal service, and we don't recommend trades or predict prices. This review covers Stocksbrew as a research tool. It is not an endorsement of any call, target, or position it publishes.

What it is

Stocksbrew describes itself as "a US-stock research product with timestamped market data, source-linked daily market analysis, stock fundamentals, Watchlist monitoring and a scoped API/MCP interface for AI agents." It's a small independent product built by one person, Tarun Tomar, rather than a funded team.

The core is a stock page. You search a US ticker and get financials, valuation, filings, earnings, analyst targets, and news in one place, with prices timestamped separately from fundamentals that refresh on a slower cadence. Around that sit sector pages, screens, an archive of daily market notes, earnings and IPO calendars, and free tools including a position size calculator.

Each stock also carries a set of Stocksbrew scores across overall, growth, quality, valuation, and momentum. On the NVDA page at the time of writing, those read 87 overall, 100 growth, 93 quality, 93 valuation, and 57 momentum. Treat those as one vendor's model, not a fact about the company.

What you get free, and what costs money

The free tier is genuinely usable for research. You can search any US stock, read the plain-English brief, and track three stocks in a watchlist. There's no card required to start.

Pro is $9 a month, with a three-day trial and a seven-day refund. It adds unlimited tracked stocks, earnings and price-level alerts, and agent access with 200 agent calls a month. It also adds the feature worth reading carefully. Stocksbrew calls it "detailed calls and zones": add, hold, or trim guidance, with buy zones, trim zones, and a target price for each name.

That last group is where a journaling reader should slow down. A call and a target are judgments, and Stocksbrew says so itself in its editorial policy. That page reads: "A thesis, score, call, or interpretation is a judgment." It also states that no model or writer can reliably predict short-term market prices. Buying the subscription for the research and ignoring the calls is a perfectly reasonable way to use it.

The agent interface is the real differentiator

Stocksbrew runs a first-party MCP endpoint with OAuth, plus a REST API with a personal token. Connect a compatible client and the agent can pull fundamentals, financial history, valuation, earnings, price history, analyst context, news, technicals, market context, and your own watchlist.

The tool surface covers what you'd expect from that description: market briefs, stock detail, stock comparison, watchlist and radar reads, and alert creation. Two design choices stand out. OAuth keeps the token out of the chat. And the documented behavior when evidence is thin is to say so, rather than fill the gap with an invented thesis.

Worth being precise about the claim, though. MCP servers for market data are not rare in 2026, and general-purpose ones exist from data vendors and as community wrappers around sites like Finviz. What's less common is a first-party MCP tied to your own saved watchlist on a $9 consumer plan. That's the narrower and more accurate version of "it works with your agent."

Stocksbrew also states plainly that it cannot place trades and is not a broker, and that prices and third-party figures can be delayed. Both are the right answers.

How it compares

The honest comparison depends on what you want the tool to do.

Against StockAnalysis.com, Stocksbrew is the lighter product. StockAnalysis Pro is $79 a year, the $6.58 a month its own page leads with, or $9.99 billed monthly. It covers 130,000+ global stocks with 10 to 40 years of financial history, which is deeper on both counts than Stocksbrew. Its published Pro page lists no API, MCP, or agent integration. If you want the widest data set for manual research, that's the stronger buy.

Against Simply Wall St, the difference is presentation and scale. Simply Wall St covers 90+ markets, with broker sync, portfolio tracking, and its own Charlie AI assistant. Its free tier caps you at five company reports a month and one portfolio. It's a far bigger product with a bigger price, and its plans page documents no MCP or agent API. Confirm current pricing on its own plans page, since the numbers render live and move.

Against the screener-first tools like Finviz and TradingView, this isn't really the same job. Those are built for scanning and charting. Stocksbrew is built for reading one company carefully and being told when something changes.

One caveat on Stocksbrew's own comparison pages, of which it publishes more than a dozen. They're vendor marketing, and the site says so directly: "stocksbrew is a party to every comparison that includes stocksbrew." That disclosure is more than most tools offer. The editorial policy also says it currently uses no affiliate links in those comparisons. Read them anyway with the conflict in mind.

Where it's weak

Depth follows attention. The directory runs to hundreds of US-listed names across every sector, and a mid-cap like Boot Barn gets the same page furniture as NVDA, with scores, ratios, analyst targets, and peers. What thins out is the context around it: in a recent week Boot Barn's page carried one news story where NVDA's carried seven. If your research lives outside the names everyone is already writing about, the page will be there; the surrounding coverage will be quieter.

It's also a one-person product. That shows up as fast iteration and a real human answering email, and it shows up as key-person risk on a tool you might build a workflow around. Fundamentals refresh on a slower cadence than prices, which is fine for position research and wrong for anything intraday.

And the scores and calls are a model, not an audit. A 93 valuation score is a compressed opinion about a company, and compressing an opinion into a number makes it feel more certain than it is.

Who it fits

It fits a position or swing trader who researches a handful of large-cap names and wants one page instead of eight tabs. It fits best if you have already started running an AI agent as part of that workflow. At $9, with a free tier that tracks three stocks, testing it costs a few minutes.

It doesn't fit someone who needs broad small-cap coverage, real-time data, or a screener as the center of the workflow. It also doesn't replace a journal. Research tells you what to consider before the trade. A journal tells you what you did and whether your process held, which is a separate record. If you are weighing journals as well, how to choose a trading journal covers that decision. What to put in a trading journal covers that side.

Frequently asked questions

Is Stocksbrew free?

There's a free tier with no card required: search any US stock, read the brief, and track three stocks. Pro is $9 a month for unlimited tracking, alerts, calls and targets, and agent access, with a three-day trial and a seven-day refund.

Does it work with Claude or ChatGPT?

Yes, through a first-party MCP endpoint using OAuth, or a REST API with a personal token. The site lists Claude Code, Codex, Cursor, Hermes, and OpenClaw as compatible clients. Pro includes 200 agent calls a month.

Can it place trades for me?

No. Stocksbrew states it is research infrastructure, cannot place trades, and has no brokerage access. It also notes that prices and third-party figures can be delayed, so treat it as research rather than an execution or real-time tool.

---

Research is the half that happens before the trade. For the half that happens after, start journaling free.