Documentation hub

Recovery Tool docs

One page that ties together how the web intake routes cases, what inputs are supported, how coverage indicators work, and when to continue in the desktop engine.

What Recovery Tool is

Structured intake plus an offline engine. Start on the web to classify inputs and route the recovery path; switch to desktop for legacy schemas, deep analysis, and execution.

Web intake (index)

Upload a backup or search by address / wallet ID / filename. The router inspects structure, flags obvious issues, and points you to the correct workflow without requesting seed phrases.

Classifies backups, addresses, and key artifacts to infer network family and script type.
Highlights supported assets, derivation hints, and whether the case is typical or conditional.
Provides direct links to downloads or coverage details when deeper work is required.

Desktop engine

Offline tooling for complex or incomplete cases. Handles legacy schemas, deep path scans, descriptor/PSBT analysis, and larger backups that shouldn’t be parsed in the browser.

Extends routing signals into execution with schema-specific handlers.
Supports air-gapped environments; run locally with no network dependency.
Preferred when exports are proprietary, partial, or involve sensitive key material.
Data handling
Never paste seed phrases. Use structured exports (backups, descriptors, xpub/PSBT, keystore JSON) and let the router guide the next step.

Intake flow

Mirrors the start page: fast routing on the web, then a clear handoff to desktop when needed.

01
Provide one input

Drag-and-drop a backup (ZIP/JSON/DAT/TXT) or search by address / wallet ID / filename. Keep original structure.

02
Detect & classify

Format checks, light parsing, and heuristic routing: family, script type, and whether the case is primary, conditional, or atypical.

03
Guided action

Shows the recommended path: stay in web for validation, open the coverage matrix entry, or proceed to desktop.

04
Desktop handoff

Offline engine continues with deeper analysis, legacy handlers, descriptor/PSBT inspection, and scan scope validation.

Links
Start at Web intake. If the router marks your case as conditional/not typical, open Coverage or download Desktop.

Coverage signals

Pulled from the coverage matrix: indicators describe routing confidence, not a guarantee of recovery.

Indicator legend

Primary Conditional Not typical
Primary — typical exports route cleanly; web intake can validate structure.
Conditional — supported with the right export (descriptor/policy/keystore) or chain context.
Not typical — classification still works, but execution usually needs the desktop workflow.

What the matrix flags

Filters by input type, script family (BIP44/49/84/86/48/custom), and wallet type. Rows include typical exports and common pitfalls.

Passphrase / 25th word, wrong purpose, or coin_type differences that change derived addresses.
Missing change chain or account index issues that make balances look empty.
Multisig completeness: cosigners, policy (m-of-n, script type), and descriptor/PSBT availability.
Export completeness: partial keystores or proprietary formats that require desktop analysis.
Deep dive
Use Coverage for filtering, details, and notes per wallet row. Pair it with Inputs & formats when choosing what to provide.

Inputs & formats

Condensed from the dedicated guide: provide one structured input at a time. The router infers network and script family.

Backups & exports

Accepted on the web: ZIP / JSON / DAT / TXT. Keep original structure; encrypted backups are fine. Router scans for wallet JSON, keystore blobs, indexes, and headers.

Address / wallet ID

Used for routing when no backup exists. Helps infer chain family and script type, and can surface coverage hints from the matrix.

Key artifacts

xpub/ypub/zpub, descriptors, PSBT, or WIF improve path inference in compatible scenarios. Multisig still requires policy + cosigners.

ZIP JSON / DAT Address / ID xpub / descriptor PSBT
Safety
Do not paste seed phrases into the web flow. If a seed is your only artifact, use the desktop guided capture or your wallet’s official restore path.
More detail
Full breakdown lives in Inputs & formats — including derivation path examples, multisig requirements, and descriptor notes.

Troubleshooting highlights

Top issues from the FAQ, mapped to the fastest corrective action.

Missing funds — often on change chain or a different account. Verify change = 0/1 and purpose (44/49/84/86).
Wrong purpose — restoring with the wrong BIP purpose produces valid but unrelated addresses.
Passphrase enabled — BIP39 passphrase alters derived keys; ensure context before scanning.
Multisig incomplete — missing cosigner xpubs or policy will block execution; PSBT helps confirm policy only.
Export incomplete / proprietary — classifiable on the web, but execution typically needs desktop handlers.
Need more
The full troubleshooting accordion remains in FAQ. Use it with Data handling if you’re unsure what to submit.

When to use desktop

Desktop is the continuation of the web intake — offline, broader handlers, and deeper analysis.

Use desktop immediately if…

Exports are proprietary or partially corrupted, cases involve large archives, or you need legacy schema support (older wallets, custom derivations).

What it adds

Full schema interpretation, descriptor/PSBT inspection, adjustable scan scope, and offline execution for sensitive key material.

Handoff model

Use the web flow to classify and set expectations, then continue in desktop with the same artifacts for deeper validation.

Download
Get installers from Desktop downloads. Keep the same backup structure you uploaded on the web.

Support & references

Quick links to the dedicated pages plus contact paths.

Coverage — filters, indicators, and wallet notes.
Inputs & formats — what to provide and how we route it.
FAQ — troubleshooting accordion with common causes and fixes.
Web intake — upload or search entry point.
Desktop tool — offline engine for advanced cases.
Contact
Email support@recovery.tool for support cases. Include the input type you tried (backup, address, descriptor) and any error notes from the web intake or desktop run.