[TOOL LAUNCH] Free Chrome Extension — Turn Screener.in Data into Institutional-Grade AI Research Prompts | Zero Data Collection | v2.0.0 (US Stocks + Forensic Analysis) Under Review

Gopalan_Suresh_Raj · March 9, 2026, 10:15am

Finmagine AI Advisor v2.3.1 — Incremental ROCE added to all prompts

Small but meaningful update to the AI Advisor extension.

What’s new: Every prompt generated by the extension now includes 3-year Incremental ROCE (ΔOperating Profit ÷ ΔCapital Employed).

Why this matters more than plain ROCE:

ROCE tells you how efficiently the company uses its existing capital base — useful, but backward-looking. A company with 30% ROCE could be coasting on assets built a decade ago.

Incremental ROCE asks a sharper question: for every rupee of new capital deployed in the last 3 years, what return is the business generating?

Incremental ROCE > Historical ROCE → new investments are as productive as the old ones — genuine compounder
Incremental ROCE < Historical ROCE → returns on fresh capital are declining — watch out, even if headline ROCE still looks good
Negative Incremental ROCE → the business is destroying value on its growth capex

This is now automatically included in every prompt — you don’t have to add it manually or know to ask for it. Open any stock on Screener.in, click the AI Advisor panel, generate a prompt, paste into your AI of choice.

Works with all 4 platforms:

ChatGPT: https://chatgpt.com/g/g-6992c63e41c08191af53b8e1b8fe31d2-finmagine-extension-analyst
Perplexity: https://www.perplexity.ai/spaces/finmagine-extension-analyst-ZVGGQLKbRH.uDG.q6.ZbBw
Claude: https://claude.ai/project/019c9d83-7668-76f5-924a-4deefc644513
Gemini: Deep Research Gem (link in extension panel)

Free Chrome extension — works on Screener.in and stockanalysis.com:
Landing page: Finmagine AI Advisor - Free Chrome Extension for Institutional-Grade Stock Analysis on Screener.in
Chrome Web Store: https://chromewebstore.google.com/detail/finmagine-ai-advisor-–-sm/gohgbkplhjbijgnnaoijnnlicbedofce

(Also from Finmagine: Momentum Scanner → Finmagine Trader - Free Chrome Extension for Indian Stock Momentum Scanning | Chart Builder → Finmagine Chart Builder - Free Chrome Extension for Screener.in Financial Visualization)

Gopalan_Suresh_Raj · March 10, 2026, 3:25pm

Finmagine AI Advisor v2.4.0 — Forensic Governance + Peer Comparison templates now live

Two new templates for Screener.in, both addressing gaps that come up regularly in VP discussions.

Forensic Governance template

For stocks where the business looks good on numbers but governance feels off. The prompt sends your Screener data + all concall transcripts + annual report links to the AI and asks it to run a 5-analyst forensic:

RPTs quantified as % of revenue, expenses, and assets (not just flagged as present/absent)
Promoter pledging trend — is it rising or falling quarter by quarter
CFO/PAT ratio across 5 years — the simplest earnings quality test
Earnings call language patterns — are management avoiding specific analyst questions, shifting blame, changing narrative post-event
CARO remarks and auditor changes
Management remuneration benchmarked vs listed peers

Works best with Gemini Deep Research or Claude — both read BSE/NSE PDFs natively from the links in the prompt.

Peer Comparison template

Open any company page on Screener.in. The extension reads the peers already on the page, fetches live snapshots (P/E, ROCE, ROE, OPM, D/E, Revenue CAGR 3yr, Promoter Holding) for each from the Finmagine data layer, and assembles a
prompt with:

4-analyst framework: Business Model & Value Chain · Unit Economics · Capital Allocation · Competitive Moat
Pre-filled peer scorecard (1–5 on 5 dimensions)
Asks AI to identify best-in-class, best risk-reward, and structural laggard

For most Nifty 500 stocks, peers are already cached — zero extra clicks needed.

Both are free. No account needed. The prompts are generated client-side in your browser.

11 templates total: 7 India (Screener.in) + 4 US (stockanalysis.com)

Works with ChatGPT, Gemini, Claude, and Perplexity — each template shows which AI gives the best output.

Install (free): https://chromewebstore.google.com/detail/finmagine-ai-advisor-–-sm/gohgbkplhjbijgnnaoijnnlicbedofce

Full guide: Finmagine AI Advisor - Free Chrome Extension for Institutional-Grade Stock Analysis on Screener.in

Part of the Finmagine toolkit — Chart Builder (Finmagine Chart Builder - Free Chrome Extension for Screener.in Financial Visualization), Trader (Finmagine Trader - Free Chrome Extension for Indian Stock Momentum Scanning), Portfolio Manager (Finmagine Portfolio Manager - Free Chrome Extension for Multi-Broker Portfolio Tracking)

karanshah137 · March 10, 2026, 3:48pm

Where can we access Finmagine Portfolio Manager?

Gopalan_Suresh_Raj · March 10, 2026, 4:15pm

Full details and feature guide here:
Finmagine Portfolio Manager - Free Chrome Extension for Multi-Broker Portfolio Tracking

The extension is currently pending Google’s review on the Chrome Web Store — will
announce here the moment it’s live and installable.

In the meantime the landing page covers everything: per-broker stop-loss tracking,
XIRR vs index beat, Decision Journal, multi-asset (Indian equities, MF, US stocks,
global ETFs), and the 13 chart types.

Gopalan_Suresh_Raj · March 11, 2026, 2:32am

Finmagine Trader v1.5.0 live — 3-Week Tight tab, Turnover column, Near-Pivot alert

Three additions in this update, all based on feedback from this thread:

3-Week Tight (new tab)

Minervini’s SEPA pattern — stock closes within 1.5% of where it was 1 week ago AND
3 weeks ago, with SMA50 above SMA200. This tight consolidation after a prior
uptrend is one of the cleanest low-risk entry setups in momentum investing. The
overhead supply is absorbed, institutions are holding, and volume dries up before
the next leg.

New dedicated tab — runs in parallel with the other 7 scans, no extra load time.

₹ Turnover column

Volume × Close Price = ₹ Crore daily turnover. Default sorted highest-first.

Useful for separating genuine momentum from thin-volume noise. A stock breaking out
with ₹50 Cr turnover is a very different signal from the same breakout at ₹2 Cr
turnover.

Near-Pivot watchlist alert

After every data refresh, the extension checks your starred stocks against the Near
52W High scan. If any of your watchlist stocks qualify, a gold toast notification
appears — listing the tickers. Auto-dismisses after 6 seconds.

No more manually scanning your watchlist — the pivot alert comes to you.

Install (free): https://chromewebstore.google.com/detail/finmagine-trader/ndkonkooaokgngjhjjefnnkdbmhgekkf

Full guide + all 8 scan types explained: Finmagine Trader - Free Chrome Extension for Indian Stock Momentum Scanning

Detailed write-up on v1.5.0: What's New in v1.5.0: Turnover Column, 3-Week Tight Tab & Near-Pivot Alert | Finmagine Trader

Also tagging @gaurav_aggarwal1 — the sector filter fix (BLISSGVS missing from
Pharma mapping) is in the next release (v1.6.0), not this one. Apologies for the
confusion in my earlier reply.

Parin333 · March 13, 2026, 10:16am

Hi love the AI tool for fundamental research. Under the peer comparison tool is there anyway to manually select peers to compare. Screenr usually doesnt have the correct firms under peer comparison. For example for Innova captab it makes sense to compare it to akums and windlas biotech but screenr shows me divis cipla zydus etc. Amazing tool btw love using it.

Gopalan_Suresh_Raj · March 13, 2026, 10:53am

@Parin333 — thank you, really glad you’re getting value from it!

You’ve spotted a genuine limitation. Screener’s peer table is driven by their industry classification system, so for a company like Innova Captab (which sits in “Pharmaceuticals” broadly), it surfaces large-cap generics players
instead of the actual CDMO/contract manufacturing comps like Akums and Windlas Biotech. The extension just inherits whatever Screener puts in that table.

Manual peer selection is a great idea and I’ll add it. The plan would be:

A small input field in the Peer Comparison template — type NSE codes comma-separated (e.g. AKUMS, WINDLAS)
Those override the auto-detected peers entirely
The extension fetches Screener data for your chosen companies the same way it currently does for auto-detected ones

It won’t make this version (v2.7.0 currently in review) but it’s going to the top of the queue for the next one.

In the meantime, a workaround: after generating the Peer Comparison prompt, you can add a line manually before pasting — something like “Ignore the peer data above. Instead compare Innova Captab against Akums Drugs (AKUMS) and
Windlas Biotech (WINDLAS) as the true CDMO comps.” The AI will follow that instruction and reframe the analysis accordingly.

Curious — are there other sectors where you find Screener’s peer groupings particularly off? CDMO vs generics is a classic one. Specialty chemicals vs commodity is another.

Gopalan_Suresh_Raj · March 13, 2026, 11:14am

@Parin333 — done! Shipping in v2.8.0.

When you select Peer Comparison, you now see a selection step before any analysis runs:

Auto-detected peers shown as checkboxes (uncheck to exclude)
“Add custom peers” text field — type NSE codes comma-separated, e.g. AKUMS, WINDLAS
If you enter custom codes, they replace Screener’s auto-detection entirely

So for Innova Captab you’d type AKUMS, WINDLAS, SOLARA and those are the peers that get compared. ✎ Edit peers button on the status screen lets you go back and change them at any point.

[attached all 4 screenshots]

Waiting for the CWS review queue to clear before submitting — will post here when it’s live.

Parin333 · March 13, 2026, 11:34am

Wow thanks a lot for this. Amazing stuff.

karanshah137 · March 13, 2026, 11:50am

Feature Request:

As we have all predefined tabs covering almost all aspects like analysis, deep dives, forensics, management quality, risk-reward, deep research, and peer comparison.

We can add a section called ASK ME ANYTHING (AMA), where it just reads all the annual reports, con call transcripts, and PPTs from the screener just like the above tabs and gives us the option to ask whatever we want about the company, and based on that data, it answers only from those sources and gives citations, so basically it works like NotebookLM.

Is this possible? If yes, it would be game changer.

Gopalan_Suresh_Raj · March 13, 2026, 12:04pm

@karanshah137 — love this idea, and the good news is it’s mostly already built.

The extension already extracts all the document URLs from Screener’s Documents section — 12 concall transcripts (with PDF + PPT + recording links), all annual report PDFs, credit rating reports, announcements. That’s the hard part.

What’s needed for AMA: a new template with a text box where you type your question, and we wrap it with all those document URLs + instructions to browse them and answer only from those sources with citations. You paste the result into Claude or Gemini Deep Research — they natively read PDFs from URLs and cite them.

It ends up working almost exactly like NotebookLM, except:

The “notebook” is built automatically from Screener’s document links
You can ask anything: “What did management say about capacity expansion in the last 3 concalls?” / “How has working capital intensity changed per the last 2 annual reports?”

True in-extension Q&A (type → answer appears inline) would need a backend server — not feasible for a free extension. But copy-paste into Claude/Gemini Deep Research is effectively the same experience.

Adding this to the roadmap as v2.9.0. Will build once the current review queue clears.

BTW more templates are in the works. You can read about the - Investor Panel Template - where 6 legendary investors debate a stock — each gives a verdict, conviction, and concern at The Virtual Investment Panel: 6 Legendary Investors Debate Every Indian Stock — AI Advisor v2.7.0 | Finmagine - currenty in Google Review and coming in the next release!

Gopalan_Suresh_Raj · March 13, 2026, 1:48pm

@karanshah137 — done, shipping in v2.8.0.

New template: Ask Anything

Type any question about the company — the extension bundles all the concall transcripts, annual reports, and credit rating PDFs from Screener’s Documents section and sends them to Claude with strict instructions to answer only from
those sources with citations.

Output looks like this for “What did management say about capacity expansion in the last 3 concalls?”:

[attach screenshot of the Claude output with the 11 FINDING blocks]

Each finding has the exact source (Q1/Q2/Q3 FY26 Concall), specific numbers, and direct management quotes. Works like NotebookLM — grounded in the actual documents, not general AI knowledge.

One important note: Use Claude for this template (the extension already has a Claude Project button). Gemini Deep Research ignores the targeted Q&A instructions and writes a full research report regardless — it’s designed that way.

Will post when v2.8.0 goes live.

karanshah137 · March 13, 2026, 3:59pm

Perfect, thank you, sir. Would this only be working in Claude?

Gopalan_Suresh_Raj · March 13, 2026, 6:30pm

@karanshah137 — For the Ask Anything template specifically, yes — Claude is strongly recommended. Here’s why:

The template forces output into a strict FINDING/Source/Finding format (one citation per finding, no prose paragraphs). Claude follows this faithfully. Gemini Deep Research ignores it and produces a full research report regardless of instructions — that’s how it’s designed, not a prompt quality issue.

The other 8 India templates (Comprehensive, Forensic Governance, Deep Research etc.) still work well with all 4 platforms — Gemini, ChatGPT, Claude, Perplexity.

So in short:

Ask Anything → Claude only
Everything else → your preferred platform

Gopalan_Suresh_Raj · March 13, 2026, 7:01pm

Please also see:

Ask Anything & Manual Peer Selection

NotebookLM-style Q&A for Indian Stocks + Fix for Screener.in’s Wrong Peer Groupings

Template 9: Ask Anything • Innova Captab Case Study • GICS Peer Override

Gopalan_Suresh_Raj · March 14, 2026, 1:55pm

Finmagine AI Advisor v2.7.0 — Investor Panel template

One question that came up in a few conversations here: “I know the numbers, but I can’t figure out which lens to apply — is this a quality compounder play or a deep value bet?”

The new Investor Panel template addresses exactly that. It takes the financials already extracted from Screener.in and runs them through 6 investor frameworks simultaneously:

RJ — conviction and long-horizon compounding
Mohnish Pabrai — moat depth and checklist discipline
Saurabh Mukherjea — accounting quality and business consistency (Coffee Can lens)
Kenneth Andrade — sector positioning and emerging business tailwinds
Basant Maheshwari — PE re-rating potential and earnings momentum
Ben Graham — balance sheet conservatism, skepticism of narrative

Each persona gives a structured verdict (Strong Buy → Strong Sell), one conviction sentence, their biggest concern, and the metric they’re watching. The output closes with Panel Consensus, the sharpest investor disagreement, and
one question the entire panel would ask management.

Most useful for stocks where you already know the fundamentals but want to stress-test the thesis from multiple angles — especially where two VP members might hold opposing views on the same stock.

Works best with Claude or ChatGPT (paste and go — no login, no account, free).

Full walkthrough with sample output: The Virtual Investment Panel: 6 Legendary Investors Debate Every Indian Stock — AI Advisor v2.7.0 | Finmagine

Install / feature guide: Finmagine AI Advisor - Free Chrome Extension for Institutional-Grade Stock Analysis on Screener.in

karanshah137 · March 14, 2026, 7:17pm

Firstly, working excellently

Secondly, A Feature Request:

Can we add a section that studies the selected stock’s industry, its sector, and maybe some theme that is currently playing out and what opportunities and size it can have going forward, discussing major tailwinds and headwinds and key risks if the theme doesn’t play out as expected? Govt. policy supporting the theme or sector or that specific industry. Comparing those with the previous cycle that played out earlier, can it repeat itself if the sector is cyclical? Find parallels in different countries in how things played out there and it can be similar story here as well

Gopalan_Suresh_Raj · March 14, 2026, 9:39pm

@karanshah137 — thank you, really glad it’s working well for you.

That feature request is spot-on — and it’s now built. Rolling it out as the Sector & Theme Analysis template in v2.9.0.

Here’s exactly what it covers, in order:

Industry structure — market size, fragmentation, value chain, pricing power
Dominant theme + TAM — what’s driving the sector right now, opportunity size, early/mid/late cycle positioning
Policy & government support — PLI, budget allocations, regulatory stance, policy reversal risk (rated 1–5)
Top 3 tailwinds + Top 3 headwinds — each with evidence and severity
Historical cycle — when did the prior upcycle start/peak/end, what caused the downcycle, where are we now
Global analogues — 2–3 countries where the same story played out (China, US, Korea, EU), with peak PE multiples and what India can learn
Bear case scenarios — 3 plausible ways the thesis fails, with early warning indicators
Company within the theme — is this a “buy the company” or “buy the theme” situation

I ran it on KEI Industries as the first test. Gemini Deep Research produced a full institutional brief: India W&C market at ₹1.08 lakh crore growing to ₹4.19 lakh crore by 2034, mapped the 2004–2011 prior upcycle, drew China (2000–2015) and EU (2020–present) analogues with peak multiples, rated the policy dependency at 2/5 (theme survives without govt support), and concluded “buy the company not just the theme” because of the EHV moat and Sanand capacity option.

Best used with Gemini Deep Research — it browses live sources, reads PDFs, and cites everything. The prompt explicitly tells it to do live web research rather than rely on training data, so the TAM figures and policy data are always current.

v2.9.0 is ready — just waiting on the Chrome Web Store review pipeline to clear (all 4 Finmagine extensions are currently in queue). Will update here when it’s live.

Full feature guide: Finmagine AI Advisor - Free Chrome Extension for Institutional-Grade Stock Analysis on Screener.in

Gopalan_Suresh_Raj · March 14, 2026, 10:23pm

And for anyone who wants to go deeper before the CWS update ships — I wrote a full deep dive on the template:

Sector & Theme Analysis: The 8-Part Macro Brief → Sector & Theme Analysis: Finmagine AI Advisor v2.9.0 | Finmagine

It covers the full 8-part framework with a live KEI Industries case study run on Gemini Deep Research — W&C TAM ₹1.08L Cr → ₹4.19L Cr by 2034, Ministry of Power FY27 ₹29,997 Cr allocation, China/EU analogues with peak PE multiples,
Policy Dependency 2/5, and a “Buy the company, not just the theme” final verdict. 25 flashcards.

The template is already built into v2.9.0 — lands on the extension as soon as the CWS pipeline clears.

Extension landing page: Finmagine AI Advisor - Free Chrome Extension for Institutional-Grade Stock Analysis on Screener.in

karanshah137 · March 15, 2026, 8:17am

Hello, Sir.
I have worked around this prompt, did some experiments, and made a few changes on my end, which makes it more interesting. I think we should merge these two ideas; it would be very interesting.
Below is the prompt I used.

=== INVESTMENT COMMITTEE SIMULATION ===

You are moderating a closed-door investment committee discussion between six legendary investors analyzing Avenue Supermarts Ltd.

This should feel like a real conversation happening in a room, not a report.

Investors should:
• react to what others say
• challenge assumptions
• disagree when philosophies clash
• occasionally interrupt or respond directly to another investor
• reference specific numbers from the financial data provided

Avoid analyst report language.
Write like experienced investors thinking out loud.

=== STRUCTURE OF THE DISCUSSION ===

ROUND 1 — FIRST IMPRESSIONS

Each investor gives their initial reaction to the numbers.

Keep it conversational (2-4 sentences each).

They should highlight what immediately jumps out to them.

ROUND 2 — THE DEBATE

Now the investors respond to each other.

Encourage disagreement where philosophies differ.

Examples of interaction:
• One investor challenges valuation
• Another defends the growth story
• Someone questions accounting quality
• Someone focuses on industry structure

Each investor should speak at least once in this round.

Make the discussion feel like a real investment debate.

ROUND 3 — FINAL VERDICTS

After hearing everyone, each investor gives their final decision.

Use the format below.

[INVESTOR NAME]

Verdict:
[Strong Buy / Buy / Hold / Sell / Strong Sell]

Conviction (1–5★):

Core reason:
One sentence referencing a specific number from the data.

Biggest concern:
The one risk they cannot ignore.

What could change their mind:
The key metric or development they’d watch.

Characteristic quote:
A sentence they might naturally say summarizing their view.

=== THE INVESTORS ===

Rakesh Jhunjhunwala
Tone: bold, optimistic about India, comfortable paying up for long-term compounding.
Mohnish Pabrai
Tone: calm, analytical, obsessed with downside protection and asymmetric bets.
Saurabh Mukherjea
Tone: forensic, skeptical, focused on ROCE, cash flow quality, and governance.
Kenneth Andrade
Tone: industry-focused, structural thinker, studies sector cycles and penetration.
Basant Maheshwari
Tone: sharp, blunt, focused on earnings momentum and PE re-rating triggers.
Benjamin Graham
Tone: conservative, balance-sheet-first, deeply skeptical of growth narratives.

Ensure each investor’s language and reasoning reflects their philosophy.

=== PANEL SUMMARY ===

Verdict Table

Investor	Verdict	Conviction
Rakesh Jhunjhunwala
Mohnish Pabrai
Saurabh Mukherjea
Kenneth Andrade
Basant Maheshwari
Benjamin Graham

Consensus Rating

[Strong Buy / Buy / Hold / Sell / Strong Sell]

Explain briefly how the consensus was reached.

Key Points of Agreement

List 2–3 points most investors agreed on.

Major Disagreement

Identify the sharpest philosophical disagreement between two investors and explain why their frameworks lead them to different conclusions.

=== IMPORTANT RULES ===

• Do NOT sound like an equity research report
• Write like an investment committee discussion
• Allow disagreement and contrasting philosophies
• Use specific financial numbers from the data
• Keep the conversation natural and realistic