Overview & Holdings

Overview

The Overview tab is the first thing you see: portfolio value, daily P&L, sector breakdown, and a market regime briefing. It’s built for a five-second read — is the portfolio up, down, and by roughly how much, before you’ve even decided to look closer.

Holdings

The Holdings tab is where the verdict engine lives. Each position gets:

• A Hold / Add / Trim / Exit-style verdict
• A confidence level and time horizon
• Scenario context — what would need to change for the verdict to flip

Rows expand on the first click (no more double-clicking out of habit) and auto-refresh keeps prices current without a manual reload.

At phone widths the filter and action controls stack into a content-sized header, so the first holding row remains reachable without scrolling through a desktop-sized spacer.

Senpai's tip

Toggle Research mode when adding a ticker to follow the idea without treating it as money you own — it stays out of P&L, allocation, largest-holding stats, briefings, and action-plan buckets until you add shares.

Research-only rows remain visible in Holdings and News, but the Overview’s invested-position count and performance tiles ignore them. If a portfolio contains only research ideas, FolioOrb says there are no invested positions instead of manufacturing a portfolio read.

Holding intelligence

Expanding a ticker surfaces deeper context:

• Peer-relative positioning within its sector
• Event flags (earnings, dividends, and other calendar items)
• Contribution breakdown — how much of your P&L this position is actually responsible for
• Insider activity — recent open-market buys and sells by the company’s own officers and directors, from SEC Form 4
• Financials over time — revenue, net margin, and diluted EPS by fiscal year, from what the company actually filed
• Your thesis — the reason you own it, in your own words
• Equity or ETF-specific detail depending on the holding type

The earnings radar carries the consensus EPS estimate for the upcoming report and how often the company has beaten that estimate over the last four quarters — so “reports in 3 days” comes with the bar it’s expected to clear. Coverage is thin for some tickers; when there’s no estimate, the date still shows and the app just doesn’t guess.

The insider activity panel counts only open-market buys and sells as conviction — the signal that someone chose to put their own money in or take it out. Option exercises, grants, gifts, and tax withholding are shown for completeness but never counted, because they aren’t decisions to buy or sell at the market. Funds and ETFs have no insiders, so they say so.

The financials panel reads SEC XBRL data straight from filings — revenue bars scaled to the window’s own peak, with net margin and EPS beside each year. A metric a company didn’t file shows an em dash, never a fabricated zero, and the revenue history stitches across the GAAP tag change most large filers made around 2018.

The Your thesis box is the one panel the app has no opinion about. Write why you bought the thing — the reason, the condition that would change your mind, whatever you’d want to hear from yourself later. Conviction is easy to articulate on a green day and very hard to reconstruct in a drawdown, which is exactly when you’ll want to read it back.

It’s yours alone: Claude never reads or writes it, it stays in your local database like everything else, and it round-trips through CSV import and export so it survives a move between machines. Editing happens inline, and a background refresh can’t clobber what you’re typing — if a save fails, your words stay on screen instead of disappearing.

Open Review → Inbox to put that thesis on an optional cadence. Mark it reviewed and choose 30, 60, 90, 180, or 365 days; FolioOrb brings it back when due. Cadence affects only the local review workflow—it never changes Portfolio math, a verdict, or what Claude sees.

None of this requires Claude. It’s all part of Local Intelligence, running deterministically against live market data. Claude, when connected, adds narration on top — not a replacement for the underlying numbers.

DCA plans and review decisions

Open the portfolio manager (press M) and choose DCA to mirror a recurring buy without connecting a brokerage. Each due interval waits in a local review bucket until you apply or skip it. Applied buys can be undone exactly only while their linked holding still contains the recorded shares and basis; later sales or edits can block reversal, leaving both the holding and DCA ledger unchanged.

Bulk apply, undo, skip, amount changes, and plan deletion use FolioOrb dialogs rather than browser-native prompts. Each dialog states what changes and what stays, keeps keyboard focus inside until you decide, cancels with Escape, and returns focus to the action you started from. A changed amount affects future intervals only. A plan cannot be deleted while it has applied buys: undo those buys first, then deletion removes its pending and skipped entries. If deletion is refused, FolioOrb shows the ledger’s reason in-app instead of silently doing nothing.

Import & export (CSV)

Open the portfolio manager (press M) and you’ll find Export CSV and Import CSV next to the holdings search.

Export downloads your active holdings — positions and watchlist alike — as a clean CSV with six columns: ticker, shares, avg_cost, is_watchlist, hold_class, notes. It opens correctly in Excel, and it is the import template, so whatever you export imports straight back in.

Import reads holdings back in, and how forgiving it is depends on whether Claude is connected — the panel tells you which mode you’re in:

• Local mode (no API key needed). Uses the exact template above: plain numbers, the six known columns, up to 200 rows. Predictable and completely offline. If your file has columns it doesn’t recognize, it says so plainly and points you at the template — it never fails silently.
• Claude assist (when a key is configured). Drop in almost any brokerage export — differing headers like Symbol or Qty, currency symbols, extra columns — and Claude maps the columns onto the FolioOrb format for you. Every row still passes the same strict checks before it’s added, so Claude only ever widens what’s accepted; it never skips validation. Clean template files skip Claude entirely (and cost nothing), and if Claude is ever slow or unavailable the import falls back to the strict local parse and tells you it did.

Either way, you get a per-row summary: what was added, what was skipped (existing holdings are skipped, never overwritten), and what errored and why. In Claude mode, Senpai adds a one-line recap on top.

Senpai's tip

Editing a CSV in Excel? Watch out for tickers like MAR26 — Excel loves to “helpfully” reformat those into dates. If a row bounces on import with a date-looking ticker, format that column as Text and re-export.