Collect years of statements and receipts. Skip the endless downloading.
PaperPull retrieves the statement and receipt history your banks, cards, stores and utilities already hold, organizes it as PDFs on your own computer, and turns your purchases and transactions into spreadsheets. It is a family of small, read-only tools that sign in alongside you: you type the password and answer the two-factor prompt in a real browser window, and PaperPull reads what you can see. Free and open source, no account, no cloud. Made for archiving (into paperless-ngx or a folder of your own) instead of clicking through each site by hand.
Watch it: a 75-second demo on the real panel with made-up accounts. Website: paperpull.net, with the getting-started guide, the provider directory and the Paperless-ngx tutorial. Download: the latest release, Windows 10 or 11 and macOS on Apple Silicon, Linux from source.
It also turns those PDFs into spreadsheets. Your receipts become one long table of every item you have ever bought, across every store and every account, with what it cost and when. The transactions inside your bank and card statements are read out of the PDFs into a workbook, one row each, and every statement is checked against its own printed balances, so a row you are looking at is a row that adds up. One button on the control panel, or one command. See Every purchase in one spreadsheet and The transactions inside your statements.
Runs on Windows and macOS (and Linux), with the same commands on each.
PaperPull is free, and it costs money to make. There is a server and domains to keep paid, developer accounts for the signed Mac build and the Microsoft Store, the tools it is built with, and evenings on every provider to build it and to keep it working as sites change. If it saves you time, you can show your appreciation and support future development by sponsoring on GitHub or donating on Ko-fi, or by buying the Microsoft Store edition for $9.99, one time, the same program with a signed installer and automatic updates. Or download it free from the Releases page. Nothing is held back from the free build. See Support.
Forty-three providers are supported today, all built on the same pattern. Thirteen more, Wells Fargo, SBA, Verizon Mobile, Golden 1, E*TRADE, State Farm, Newrez, Kroger, Meijer, American Family, Apple Card, Stripe and FedEx, are built and waiting for someone with an account to test them (#27, #28, #31, #35, #36, #37, #38, #41, #42, #45, #52):
| App | Provider | Documents | Notes |
|---|---|---|---|
affirm |
Affirm | Loan agreements (Truth in Lending), one per loan | JSON API from inside the page, nothing clicked; no monthly statement exists for a pay-over-time account |
adp |
ADP Workforce Now | Pay statements, W-2s | Statement services from inside the page, nothing clicked, real Edge/Chrome. Requested in #46 |
aafmaa |
AAFMAA (Armed Forces Mutual) | Annual statements, policy docs | ASP.NET WebForms; one documented disclosure dialog |
ally |
Ally Bank | Account statements, tax forms | JSON API; same-dated statements named from the PDF |
amazon |
Amazon (any country's store, marketplace setting) |
Order invoices (full history) | Per-year order pagination |
amfam |
American Family Insurance | Billing statements, policy documents, declarations, ID cards | Untested, built without an account. Have one? Help test it Being tested by @jpfieber. |
amex |
American Express | Statements, Year-End Summary | Click-nav SPA; in-memory session |
att |
AT&T (Mobility, Fiber, Internet) | Monthly bills | Pilot confirmed by @watling777 in round eight. The full run and a second account are the rounds still open (#26) |
anthem |
Anthem BCBS (Elevance, 14 Blue states) | EOBs, plan docs (all years), ID cards, letters | Health insurance (PHI); tRPC API, nothing clicked. Contributed by @riordan |
applecard |
Apple Card and Apple Savings | Apple Card statements, Savings statements, tax forms | Untested, built without an account. Have one? Help test it Being tested by @watling777. Requested in #52 |
bestbuy |
Best Buy | Online, store-order and in-store receipts, and returns | Built on the maintainer's own account, 35 purchases back to 2015. The history through the page's own query a year at a time, each details page saved as the receipt, nothing clicked, real Edge/Chrome, paced for Best Buy's bot protection |
capitalone |
Capital One | Bank and card statements, tax forms, letters | Ported by @davidrudnick; fresh live pilot pending |
chase |
Chase (credit cards) | Card statements | Real Edge/Chrome; per-card accordions + year picker |
citi |
Citi (credit cards) | Monthly card statements | Statements API from inside the page, nothing clicked; the site lists about two years online |
costco |
Costco Wholesale | Receipts, in-warehouse, gas, pharmacy and costco.com orders | Written from a member's recording and run against their account, 34 purchases back to July 2024 (#47). One membership so far, so a second Pilot is welcome. Two tabs, quarters walked, the warehouse receipt is a dialog rendered with printToPDF, real Edge/Chrome |
discovercard |
Discover (credit cards) | Card statements | Capital One is moving these accounts onto its own site. Once yours has moved this app can no longer read it (#13) |
dominion |
Dominion Energy (VA) | Billing statements | Paginated MUI accordion; ~18-month limit |
ebay |
eBay | Order receipts, ten years of purchase history | Confirmed by @jpfieber, every order back to 2017. Order-details page rendered to PDF, nothing clicked, real Edge/Chrome. Requested in #44 |
lowes |
Lowe's Home Improvement | Store and online receipts, and returns | Built on the maintainer's own account. The purchase history read by address, each purchase's details page saved as the receipt, nothing clicked, real Edge/Chrome |
homedepot |
The Home Depot | Order receipts, with model and store SKU numbers | Built on the maintainer's own account. The history through the page's own request, Home Depot's own print receipt taken without opening the print dialog, nothing clicked, real Edge/Chrome. Home Depot keeps two years online |
etrade |
E*TRADE | Statements, trade confirmations, tax forms | Untested, built without an account. Have one? Help test it Being tested by @watling777. |
ezpassva |
E-ZPass Virginia | Monthly and quarterly toll statements | Built on the maintainer's own account. The portal's own statements list read from inside the page, nothing clicked; about a year is online, so run it at least once a year |
fairfaxwater |
Fairfax Water (VA) | Water bills | Mendix portal, driven like a person; PDFs for the last year only, so run it quarterly |
fedex |
FedEx Billing Online | Shipping invoices | Untested, built from a login not connected to Billing Online. Use it? Help test it Requested in #54 |
fidelity |
Fidelity Investments | Statements, trade confirmations, tax forms | Document Access Hub API from inside the page, nothing clicked; real Edge/Chrome |
gap |
Gap Inc. (Gap, Old Navy, Banana Republic, Athleta) | Order receipts | Lazy-loading history; ~13-month limit |
kroger |
Kroger (Pick 'n Save, Metro Market, Fred Meyer, Ralphs and the rest) | Receipts, in-store, fuel, pickup and delivery | Built against an account with no purchases yet. Have one? Help test it Being tested by @jpfieber. Purchase-history API and receipt page, nothing clicked, real Edge/Chrome |
github |
GitHub | Payment receipts (Pro, Copilot, Actions, Sponsors and the rest) | Confirmed by @jpfieber on #43. Receipt links fetched or printed, nothing clicked |
golden1 |
Golden 1 Credit Union | Statements, tax forms | Untested, built without an account. Have one? Help test it Being tested by @watling777. |
meijer |
Meijer | Order receipts, in-store digital receipts where mPerks lists them | Untested, built without an account. Have one? Help test it Being tested by @jpfieber. Nothing clicked |
myecp |
MILITARY STAR (Exchange Credit Program) | Card statements | Built on the maintainer's own account, 25 statements back to 2023. The MyECP statements list and download from inside the page, nothing clicked |
mypay |
DFAS myPay | eRAS, CRSC, 1099-R, 1095 | Government pay system; JSON API, nothing clicked |
mtb |
M&T Bank | Mortgage statements, escrow, 1098 | Own online banking; you list, app expands all years |
netbenefits |
Fidelity NetBenefits (workplace 401(k)) | Quarterly or monthly statements, made to order | The site generates statements on request; the app requests each period and renders it, nothing clicked |
navyfederal |
Navy Federal CU | Account statements | Per-account accordions; blob-tab PDFs |
newrez |
Newrez | Mortgage statements, escrow analysis, 1098 | Untested, built without an account. Have a Newrez mortgage? Help test it Being tested by @watling777. |
paypal |
PayPal | Monthly statements | Built on the maintainer's own account, 25 statements. The site's own statements list and download from inside the page, nothing clicked; three years are online |
paylocity |
Paylocity | Pay statements | Escher JSON API, enqueue-poll-fetch PDF; nothing clicked |
pge |
PG&E (Pacific Gas and Electric) | Billing statements | Salesforce portal with a paginated history. Contributed by @appchamp, a pagination repair being tested by @watling777 |
redcard |
Target RedCard / Circle Card (TD Bank) | Billing statements | Statements table; per-year switcher |
robinhood |
Robinhood | Account statements, tax docs | "View More" pagination |
sba |
SBA (MySBA Loan Portal) | Loan statements, 1098 | Untested, built without an account. Have an SBA loan? Help test it |
schwab |
Charles Schwab | Statements, tax forms, letters, trade confirmations | Ported by @davidrudnick; fresh live pilot pending |
smud |
SMUD (Sacramento Municipal Utility District) | Monthly bills | Untested, built without an account. Have one? Help test it Being tested by @watling777. |
stripe |
Stripe (merchant Dashboard) | Fee invoices, tax forms such as the 1099-K | Untested, built on an account with no documents yet. Have Stripe invoices or tax forms? Help test it Requested in #53 |
statefarm |
State Farm | Bills, renewal notices, ID cards, receipts, policy documents | Untested, built without an account. Have a policy? Help test it Being tested by @watling777. |
target |
Target | Receipts (Online + In-Store) | Print-capture |
tmobile |
T-Mobile | Bill statements | Bill-history page; detailed-bill download |
tsp |
Thrift Savings Plan | Participant statements, 1099-R | Secure Mailbox API from inside the page, nothing clicked; downloading marks the message read |
ukg |
UKG Pro / UltiPro | Pay statements | Per-employer tenant; JSON-API, nothing clicked |
ups |
UPS Billing Center | Shipping invoices | Built on the maintainer's own account. The Billing Center's own invoice list and download from inside the page, nothing clicked |
usaa |
USAA | Statements | JSON-API enumeration |
usbank |
U.S. Bank | Credit-card statements | Ported by @davidrudnick; fresh live pilot pending |
verizon |
Verizon (Fios) | Bill statements | Real Edge (bot block); dropdown + CDP download |
verizonmobile |
Verizon Mobile (wireless) | Monthly bills | Untested, built without an account. Have one? Help test it |
walmart |
Walmart | Receipts | Hardened against bot detection |
wellsfargo |
Wells Fargo | Account statements, tax documents | Untested, built without an account. Have one? Help test it |
wealthfront |
Wealthfront | Statements, tax docs |
⚠️ Read this first: these tools drive real, signed-in financial accounts. See SECURITY.md before you run or publish anything. In short: never commit your*-browser-profile/folder, yourconfig.json, or any downloaded PDF. The.gitignoreblocks them, don't override it.
Your documents live on the provider's site, and it will only hand them to a browser that is already signed in. So PaperPull never tries to be you, it works beside you. You sign in yourself, in a real browser window, and the tool attaches to that window afterwards and reads.
flowchart TB
you(["You"]) -->|"sign in · 2FA · device approval"| br["A real browser window<br/>its own profile · its own debugging port"]
br -.->|"attaches over CDP, reads, never authenticates"| app
subgraph app ["One app = one provider"]
orch["Orchestrator<br/>discover → download → verify<br/>the same in every app"]
site["provider_site.py<br/>selectors · URLs · download quirks"]
core["paperpull-core<br/>naming · filing · state · CSV · PDF checks"]
orch --> site
orch --> core
end
app --> out[("Your folders<br/>PDFs + an index CSV")]
That single choice is why there is no password anywhere in this project, why 2FA and device approvals are never an obstacle, and why a provider tightening its login breaks nothing here.
In practice that first step is paperpull <app> login (or the app's own
login.bat / login.command), which opens the browser for you, a plain
Chromium for most apps, or your own installed Edge/Chrome for the few sites
whose bot detection turns a fresh Chromium away (Walmart, Verizon, Chase).
Each app gets its own profile and its own debugging port, so several
signed-in browsers can sit open at once without colliding. Those profiles
are separate from your everyday browser on purpose, so each starts with no
extensions and no saved logins. They are real browser profiles, though, so
a password manager installed into one, from the extension store in the
window Login opens, stays there for every Login of that provider after.
It is one profile per provider, so the extension goes in once for each
provider you set up, not once for all of them.
Everything a provider knows lives in one file. provider_site.py holds
every selector, URL and download quirk for that site. The orchestrator around
it is the same in every app, and paperpull-core underneath it is
shared. When a provider redesigns, the repair is one file, never a rewrite,
and never a change to how documents get named, filed or tracked.
flowchart TB
D["Discover<br/>list what the provider still has"] --> Q{"Already downloaded?"}
Q -->|yes| S["Skip it"]
Q -->|no| DL["Download the PDF"]
DL --> V{"Is it a real PDF?"}
V -->|no| MR["Manual Review<br/>flagged, never silently lost"]
V -->|yes| F["Classify, name, file<br/>+ append to the index CSV"]
F --> OK["Mark downloaded_ok<br/>sticky, survives deletion"]
Three plain-text files carry the state, and you can read all of them:
| File | Holds |
|---|---|
discovery.json |
what the provider showed us this run |
progress.json |
what happened to each document, including the sticky downloaded_ok |
<Provider> Document Index.csv |
one row per saved document, for humans and spreadsheets (receipt apps also keep an Order History.csv, one row per line item) |
That last step is what makes a re-run safe. downloaded_ok is keyed to the
document, not to the file on disk, so you can import everything into
paperless-ngx, delete the PDFs, and the next run still skips them. It only
fetches what is genuinely new, and lists it in new-this-run.txt.
Nothing that buys, sells, transfers, pays, deletes, or changes a setting is
ever clicked, and all site interaction lives in provider_site.py where it can
be read in one sitting. Every app that clicks enforces this deny-by-default, a
control must clear a blocklist (FORBIDDEN_CONTROL_RE) and match a document
allowlist (SAFE_DOC_CONTROL_RE), and the app's host allowlist refuses any
stored URL that points elsewhere. Sixteen apps click nothing at all (ADP, Affirm, Amazon,
Anthem, Citi, eBay, Fidelity, Gap, GitHub, Kroger, Meijer, myPay, NetBenefits, Paylocity, TSP, UKG), they read a JSON API or render a
page they navigated to. A repo-wide test checks every app's guard.
SECURITY.md spells out which app does which.
In the control panel, add a person beside the Account box makes the
second person's account, and the Account box then runs every action against
it. From a terminal it is paperpull <app> add-account spouse, then
paperpull <app> all --account spouse. Either way it gets its own profile,
port and output folder beside the first one's, so no data mixes. Underneath it is a
config.spouse.json beside the app's config.json, which the app also takes
directly as --config, and the sign-in launcher takes the label too
(login.bat spouse / ./login.command spouse). python tools/add_account.py spouse does every app at once.
Want to see it first? The panel's welcome screen offers See a sample archive, a folder of invented statements and receipts. The Status tab fills in and both spreadsheets build from it, so you can see exactly what PaperPull produces before you sign in to anything. Nothing downloads while you are in it, and Leave the sample puts you back. Every document in it says on its face that it is invented.
One-shot setup (creates a venv for every app + the GUI, installs the browser):
setup-all.bat REM Windows./setup-all.command # macOS / LinuxThen either drive everything from the GUI control panel, pick an app and account, click an action, and watch the live output:
gui\run_gui.bat
…or run a single app from the terminal, with one command for all of them
(using amex as the example):
copy apps\amex\config.example.json apps\amex\config.json REM then edit paths as needed
paperpull amex login REM opens a browser, sign in yourself, leave it OPEN
paperpull amex pilot REM download the newest few as a test
paperpull amex all REM download everything available
paperpull amex resume REM continue after an interruption
paperpull list REM every app it can seepaperpull is paperpull.bat on Windows and ./paperpull on macOS and
Linux, or python paperpull.py anywhere. It finds the app by folder name,
slug or provider, runs it under its own environment, and passes anything else
straight through, so paperpull chase all --year 2025 --account spouse works.
The commands are setup, login, discover, pilot, all, resume,
verify, diagnose and dry-run. The panel offers the six of those a
person uses day to day, plus a Scope row (one year, or a date range) that
becomes the same --year, --start-date and --end-date every app takes.
Each app also has its own README with provider-specific details and quirks.
(Prefer to set apps up one at a time? paperpull <app> setup, or the app's
own setup.bat / setup.command.)
tools/status.py reads the state each app already keeps and reports how current
every archive is. Copy it and status.bat next to your install folders and run
it. It downloads nothing and changes nothing.
PROVIDER DOCS NEWEST AGE ISSUES STATUS
Some Payroll 12 2026-06-18 72 d 2x month !! OVERDUE
A Bank 97 2026-06-30 60 d monthly * due
A Mortgage 93 2026-07-31 29 d monthly current
A Shop 667 2026-08-13 16 d - ongoing
It reports the archives you have. Nobody holds an account with every provider, so a folder you never set up, or one left behind by a closed account, is left out entirely rather than listed as missing or overdue. An archive that was set up but never downloaded anything is called out by name, since that is the one state a single run fixes.
It answers "is something new probably waiting" rather than "when did I last run this", which are different questions. A run that only verified existing files still updates a timestamp while telling you nothing about whether a new statement exists. So the signal is the date of the newest document you actually hold, measured against how often that provider issues them.
The cadence comes from your own history and is measured per account, so nothing has to be configured, and a provider that moves from monthly to quarterly corrects itself. Measuring per account matters: one bank folder can cover several accounts, and pooling their dates makes a monthly cycle look weekly. Receipt archives are shown without a due date, because purchases arrive irregularly and "40 days overdue" would be noise.
Being up to date is not the same as being complete. An archive can hold a document from last week and still be missing whole years behind it, which is exactly what happened twice while building this: one mortgage archive held a single year of a seven year history, and a payroll archive quietly defaulted to year to date. Both looked healthy by their newest document.
So each series is also checked for periods that look missing from the middle:
Possible gaps. A run that looks current can still be missing
periods in the middle, so these are worth a look.
A Bank
2025-03-18 to 2025-05-17 1 missing 3 series, including Savings
2026-01-01 to 2026-04-02 2 missing 3 series, including Savings
The hard part is not finding gaps, it is not inventing them. Plenty of real documents arrive irregularly, insurance ID cards and policy renewals among them, where a long quiet stretch means nothing was issued rather than something was missed. Flagging those would train you to ignore the report, so a series has to earn an opinion first: at least six documents, a median interval of ten days or more, and at least 65 percent of its intervals close to that median. Only then is an interval roughly twice the usual one reported, and it is reported as possible rather than certain.
Windows shared by several series are grouped, because one account missing a month is usually a quiet month, while the same window missing across several at once is what a run that failed part way looks like.
--html also writes a status.html dashboard you can bookmark, and --quiet
prints only what needs attention. The dashboard reads no document contents and
carries no amounts or account numbers, but it does list which providers you
hold accounts with, so it belongs with your installs and is gitignored here.
The receipt apps (Amazon, Target, Walmart, Gap) record every line item they
see while downloading, in an <Provider> Order History.csv beside the PDFs.
tools/export_purchases.py gathers all of them into one workbook, one row
per item across every provider and account, newest first, with an Orders
sheet and a Summary of spend per provider per year. Amounts are numbers, so
Excel can sum them. Nothing reads a PDF and it takes about a second.
python tools/export_purchases.py --root "C:\path\to\your\installs"
The panel has the same thing on its Spreadsheet tab, one button, with a
dropdown for one provider at a time (Amazon Purchases.xlsx). The file is
rebuilt from scratch each time, so edit a copy, not the original.
A statement archive holds PDFs, and the transactions are inside them.
tools/export_transactions.py opens each PDF a statement archive's index
knows about, reads it line by line, and keeps the lines that have the shape
of a transaction, a date, a description and an amount. Nothing in it is
written for one bank. What makes it trustworthy is the statement itself.
Every statement prints a beginning and an ending balance, and the
transactions between them have to add up.
python tools/export_transactions.py --root "C:\path\to\your\installs"
Where the statement carries a running balance column, the sign of every amount is read off the balance and the statement reconciles to the cent by construction. Where it prints signed amounts, they are summed as printed and checked against every balance pair the statement shows, which is how a card statement that prints its summary three times is read right. A statement covering two accounts is reconciled one account at a time. A statement that does not add up is still exported, with the difference in the Statements sheet, so you know which rows to doubt. On the author's archive, every USAA and Navy Federal statement and 120 of 128 American Express statements reconcile. Brokerage statements never will, since their balances include market movement, and the sheet says so.
Amounts are the effect on the balance. Money in is positive, money out is negative, for a bank account and a card alike. Each PDF is read once and remembered in a cache beside the installs, so the first build of a big archive takes minutes and the next takes seconds. The panel's Spreadsheet tab has this too, under Statements, streaming its progress.
Files are named 2026-03-09 Target Coffee Receipt.pdf unless you say
otherwise. The panel's File names tab builds your own pattern from
parts, the date in the shape you like, the provider, the order or
document number, the account, the total and more, each shown with how
many of your own files carry it. A part can be skipped when a document
does not have it, so no name ends in a stray dash, and the three newest
files you already have are shown under the new pattern as you type.
One pattern covers every receipts app or every statements app, or any single app can have its own. After you save, the tab offers to rename the files you already have to match, a preview first and then the rename, and nothing is downloaded again. The pattern language is in docs/file-naming.md.
Both have a package. Every release on the
Releases page carries
PaperPull-<version>-setup.exe for Windows, which installs the control
panel, the shared core and every provider into your own user folder with no
admin rights and no Python on the machine, and PaperPull-<version>.zip, the
same folder for anyone who would rather not run an installer. The package
is x64 and runs on Windows on ARM (a Snapdragon laptop, say) under the
emulation Windows 11 provides, where the browser it drives is your own
native Edge or Chrome and the Python side spends its life waiting on it. A
native ARM64 build is one flag away in packaging/build_windows.py and
will be shipped when someone needs it.
Two Windows editions, one program. The GitHub release is free. The
Microsoft Store edition is the same package from the same build, for
$9.99 one time, and what the price buys is convenience, a Store-signed
package that installs with no warning, updates through the Store, and
uninstalls cleanly, plus the knowledge that it keeps the project going.
Nothing is held back from either edition. The AGPL permits selling copies and the source stays public. For macOS
there is PaperPull-<version>-arm64.dmg, Apple Silicon only, signed and
notarized, so it opens with no warning. Drag PaperPull to Applications and
double-click it. The Windows installer is not yet code-signed, so Windows
shows its SmartScreen prompt the first time. See
Code signing policy below.
For a checkout of this repository, one download covers both. The two
double-click files each app keeps, and the one-shot setup, come in both
flavors, and everything else is the same paperpull command on either:
| Task | Windows | macOS / Linux |
|---|---|---|
| One-shot setup | setup-all.bat |
./setup-all.command |
| Set up one app | setup.bat |
./setup.command |
| Sign in | login.bat |
./login.command |
| Test run | paperpull amex pilot |
./paperpull amex pilot |
| Full run | paperpull amex all |
./paperpull amex all |
| Control panel | gui\run_gui.bat |
gui/run_gui.command |
A second account is the same on both: paperpull amex all --account spouse.
Only one thing genuinely differs. macOS keeps Playwright's browser inside an
app bundle and in a different cache directory, and a couple of providers need
a branded Edge/Chrome to get past their bot protection, that lookup lives in
paperpull_core.browser and is handled for you.
To use PaperPull, the .dmg above is the way. This is for a checkout of the
repository.
git clone is the smoothest route, it preserves the scripts' executable
bit and macOS does not quarantine it.
If you download a release archive instead, prefer the .tar.gz: it keeps
the executable bit, while a .zip drops it. After unpacking a download,
macOS may also quarantine the scripts, so a double-click reports "cannot be
opened because it is from an unidentified developer." Both are cleared in one
go:
xattr -dr com.apple.quarantine .
chmod +x setup-all.command apps/*/*.command gui/*.command- Windows, macOS, or Linux
- For the Windows installer or the macOS
.dmg, nothing else. They carry their own Python. - For a checkout, Python 3.11+ and Playwright (installed per app by the setup script)
- A Chromium-family browser for the few providers that need a real one (Chrome, Edge, Brave, Vivaldi or Opera). Safari and Firefox cannot be driven this way.
No one has accounts everywhere, so PaperPull grows when people add the providers they use. If a bank, card, brokerage, utility, telecom, or retailer you use isn't here yet, you're the ideal person to add it:
- 📖 Adding a provider, a step-by-step guide (clone the closest app, rewrite one file, stay read-only, test, submit).
- 📋 PROVIDERS.md, what's supported and what's requested; claim one so nobody builds it twice.
- 📥 Can't build it yourself? Request a provider and someone with that account may pick it up.
- 🎬 Testing a provider, for people who have the account but do not write code. Sign in, press Record, click your way to one document, and send back what the app wrote down. That is usually all it takes to turn a scaffold into a working app.
Every contribution keeps the read-only, local, no-credentials design, see CONTRIBUTING.md and SECURITY.md.
- ✅ All fifty-eight apps pass their tests, the forty-three that are supported and the thirteen still waiting for a tester, 8,763 of them across the repo. Most are in regular use by the author. The rest were built or confirmed by people who hold those accounts, named in the table above and in Thanks.
- ✅ Packaged. A Windows installer and a signed, notarized macOS app, both built by GitHub Actions from the tagged commit, with checksums. A Microsoft Store listing and free open-source code signing for Windows are in progress.
- 🔜 More providers: community-driven, see PROVIDERS.md.
- 🔜 Scheduled/assisted runs: a monthly "nudge + sweep" (e.g. the 1st) that opens the login browsers and then runs discover + resume across every app once you've signed in, delete-safe, so it only grabs what's new. Fully unattended runs stay out of scope by design: the tools never store credentials or bypass 2FA, so a human sign-in stays in the loop (long-session retailer apps may tolerate more automation than banks/cards).
- ✅ Shared core: the support code the apps used to duplicate now lives once
in
core/aspaperpull-core. An app declares anAppSpec, its folders, routing, CSV columns and config defaults, and keeps only its orchestrator and its*_site.py.tools/check_installs.pyreports whether your installs have drifted from the repo.
The Windows installer is built from a tagged commit of this repository by the Windows package workflow on a clean GitHub runner, never from a developer's machine, and the checksums of what it produced are published beside it. Signing goes through that same workflow so a signed binary can only ever come from code that is in this repository.
This program does not transfer any information to other networked systems unless specifically requested by the user. The only sites it contacts are the providers you sign in to yourself, and the only download it ever offers is a browser, at sign-in, with your agreement. The full statement is PRIVACY.md.
Team roles, current status and the full policy are in docs/code-signing.md.
PaperPull is free and open source, and it costs real money and real time to develop. There is a server and domain names to keep paid, developer accounts for the signed and notarized Mac build and for the Microsoft Store listing, and the tools it is built with. Every one of the forty-odd providers took evenings to build, and each one needs repairing when its site changes, which they do.
If PaperPull saves you time, there are three ways to show your appreciation and support future development. All are optional, and none changes what the free build does.
- Sponsor on GitHub: github.com/sponsors/rheeloaded, one-time or monthly, from a card you already have on GitHub.
- Donate on Ko-fi: ko-fi.com/rheeloaded ☕
- Buy the Microsoft Store edition for $9.99, one time, the same program from the same build, with a signed installer that opens with no warning and updates through the Store. The listing is in review and will be linked here once it is live. It unlocks nothing the free build lacks.
You can also help without spending anything: test a provider you hold an account with (see PROVIDERS.md), report what breaks, or contribute one. That is worth as much as a donation.
PaperPull only reaches the providers people bring to it. These people built one, tested one against an account the author does not hold, or found a bug and diagnosed it to the line.
- @davidrudnick, the Capital One, U.S. Bank and Charles Schwab apps, and a long run of panel fixes.
- @riordan, the Anthem BCBS app, the first health insurer.
- @appchamp, the PG&E app.
- @marecabo, Amazon's legal invoice PDFs on the German store.
- @watling777, the tester behind AT&T, SMUD, Golden 1, E*TRADE, State Farm and Newrez, seven providers' worth of surveys and pilots in a weekend, and the PG&E pagination bug.
- @dertbv, three exporter and Navy Federal bugs, each diagnosed so exactly there was nothing to add.
- @liamrotheram and @OberstK, Amazon outside the United States.
- @jpfieber, who asked for eBay and became PaperPull's first sponsor before it was even built, then tested eBay and GitHub to working, eBay through every order back to 2017, and is testing Kroger, Meijer, American Family and Target.
If you tested a provider and are not here, say so on the issue and you will be.
This project is for personal archival of your own records. It is not affiliated with, endorsed by, or sponsored by any of the companies listed. All product names and trademarks are the property of their respective owners. Automating access to a website may be restricted by that site's Terms of Service, you are responsible for how you use these tools. Provided as-is, without warranty of any kind.
License. PaperPull is free software under the GNU Affero General Public License, version 3. You can run it, read it, change it and share it. If you distribute a changed version, or run one as a service for other people, the same license applies to what you distribute, source included. Contributions before 2026-09-18 were made under MIT and that permission is kept, see NOTICE.md.
Name. The PaperPull name is not part of the license. A modified version needs its own name, so that anything called PaperPull is this project. What that allows and what it doesn't is in TRADEMARK.md.