Before you start

• Give the item a clear title and category. Add the make, model or edition, ISBN or barcode, size and any useful notes. For a book, the author and edition can make a big difference.
• Choose the country where you would buy the replacement, and the currency you want. Check these rather than relying on your phone or browser language.
• Check the warranty date and your insurance settings. If the warranty is still active, ThingzSafe looks for the price of a new item. If not, it looks for a used one. If you have switched on New for old, it always looks for a new one.
• You need an internet connection and a connected Pro household. Fair usage limits apply.

Value one item on Android

1. Open the Thing, Part or Content and choose Edit details.
2. Beside Current value (each), tap AI. Check the identity details in the edit form first.
3. Check Country and Currency, then tap Generate AI valuation.
4. Keep the valuation screen open while it works. Read the result, and open the VIEW links if you want to check a listing.
5. Choose a price, then tap Accept or Edit and accept. The chosen value goes back to the edit form. Save the edit form to keep it.

The magnifying glass is a separate, ordinary web search. It does not create or accept an AI valuation.

Value several items together

On Android

Open AI valuations from Needs review, or from another supported item list. Tick only the items you want, then choose Generate. If you start from Needs review, only the items in that review list are included.

On the web

1. Open Bulk Review Items.
2. Tick the AI valuation column for each row you want. The box in the heading ticks or clears all of them.
3. Choose Generate AI valuations for selected items · Pro. If you have unsaved changes in the grid, confirm the save first.
4. In the valuation window, check Country and Currency, then choose Generate.

When it finishes, the list shows just the items from that run. Choose Show all items to go back to the full list, where it is offered.

Reading the result

• The proposed value is the replacement value for one item. Next to it you will see whether it is based on a new or used item, how confident the match is, the price found, delivery, and the usual price range without delivery.
• Up to three price choices show a low, a middle and a high example. ThingzSafe starts with the higher reasonable one. Choose another and the amount updates to include that listing’s delivery. Check the Accept amount before you accept.
• VIEW opens the original listing. Listings can change or end, show a different version, or have delivery limits. Check the exact model or edition, condition, quantity, currency and postage yourself.
• Confidence shows how well the listing seems to match your item. It is not a guarantee. The range shows the prices that were found. It is not a formal valuation certificate.

Country and delivery

ThingzSafe looks at sellers in your chosen country first. It uses overseas listings only when it cannot find reliable local ones. For overseas listings it allows extra for postage, at least three times the listed delivery cost, and it never assumes missing or zero postage is free.

Delivery is an estimate, especially for large items or international shipping. Check the destination and the seller’s delivery terms before you accept.

Accept, edit or reject

• Accept uses the amount shown for that proposal.
• Edit and accept lets you type your own reviewed replacement value, including delivery.
• Reject leaves the current value as it is.
• Mark item as reviewed when accepted also clears the item’s Needs review flag when you accept its value. It is separate from the box you ticked to choose the item.
• Accept selected reviewed proposals applies the proposals you have selected and closes the valuation screen. On the web, the values go back into the Bulk Review grid. The accept button also appears below the results, where shown.

Accepted results fold away so you can carry on down the list. Your existing values do not change until you accept. Values are per item, and your inventory totals take quantity into account.

If you started from a single item’s edit form, accepting returns the value to that form, so save the form to keep it. When you accept several items together, their values are saved straight away.

If it stops or goes wrong

• Cancel remaining stops any further items once the current one finishes. Results already found are kept for you to review. Cancel closes the whole process, and values you have already accepted stay saved.
• If the connection drops, follow the message on screen. A failed lookup may need a fresh try, but one that is still working can be recovered. Please do not keep pressing Generate just because a result has not appeared yet.
• If Android offers Retry after a display problem, it only shows the result again, without another AI lookup. Use Details or Copy details when you contact us.
• If nothing reliable is found, improve the title, category, make, model or edition, or type in a value yourself.

API quick start and access tokens

Base URL and authentication

Use HTTPS at https://app.thingzsafe.com. Send a developer token in the request header:

Authorization: Bearer YOUR_TOKEN
Content-Type: application/json

Use a developer token created in the web app, not an Android/device sign-in token or an AI assistant OAuth token. Each token is tied to one household and its verified owner. Active Household or Pro cloud access is required.

Create a token

1. Sign in to the web app as the household owner.
2. Open How to use API.
3. Enter a connection name and an expiry of 1–90 days. The default is 30 days.
4. Select the required permissions and choose Create token.
5. Copy the token while it is shown. It cannot be displayed again later. Store it in your application's secret configuration.

The initial selections are inventory reading and valuation reports. Inventory changes and file reading are separate optional selections. A household can have up to 20 active, unexpired developer tokens. Use Revoke to stop an unused token; creating a replacement does not automatically revoke the previous one.

Developer write routes require inventory.write; include inventory.read as well if your application looks up records before proposing changes. AI assistant write tools require both.

Try a read in PowerShell

The following prompts for the token locally without displaying it. It uses the existing Sites in your household and changes no inventory.

$token = [System.Net.NetworkCredential]::new('', (Read-Host 'Developer token' -AsSecureString)).Password
$headers = @{ Authorization = "Bearer $token" }
$body = @{ kind = 'Site'; offset = 0; limit = 100 } | ConvertTo-Json
Invoke-RestMethod -Method Post `
  -Uri 'https://app.thingzsafe.com/v1/developer/query' `
  -Headers $headers -ContentType 'application/json' -Body $body

Try a read with curl

Put your token in a local secret/environment variable named THINGZSAFE_TOKEN. Save the JSON below as sites.json:

{"kind":"Site","offset":0,"limit":100}

curl 'https://app.thingzsafe.com/v1/developer/query' \
  -H "Authorization: Bearer $THINGZSAFE_TOKEN" \
  -H 'Content-Type: application/json' \
  --data-binary @sites.json

On Windows PowerShell, use the PowerShell example above, or invoke curl.exe with Windows shell syntax. Never publish a real token in an example, screenshot or support conversation.

Shared request format, IDs and pagination

Endpoint map

These are the developer-facing routes documented here. The web app's own authenticated management routes are not interchangeable with developer-token routes. In particular, a developer token cannot create or enlarge its own permissions.

Query fields

{
  "text": null,
  "siteId": null,
  "areaId": null,
  "categoryId": null,
  "kind": null,
  "relationship": null,
  "parentThingId": null,
  "includeDescendants": false,
  "missingValueOnly": false,
  "offset": 0,
  "limit": 100,
  "report": false
}

Do not combine a metadata kind lookup with hierarchy fields, or use a metadata kind with report:true. Use IDs, not location names, in the ID fields. Ambiguous names must be resolved before choosing an ID.

IDs may be compact 32-character GUID strings or standard hyphenated GUID strings. Keep returned IDs unchanged. All IDs in the examples below are fictional placeholders: look up and substitute your own.

Follow every page

Read totalCount, returnedCount, hasMore and nextOffset. While hasMore is true, repeat the same filters using the returned next offset. Do not report a page's item count as the total count.

Results are live rather than a frozen export. If inventory changes during paging, re-run the query for a consistent fresh assessment.

Sites

Permission: inventory.read

Use the shared query endpoint to discover Site IDs. A Site is a top-level location, such as House or Barn.

{"kind":"Site","text":"House","offset":0,"limit":100}

Read each returned item's id, kind, name and revision. Other generic fields may be present and need not be used for a Site lookup. Follow pagination if there are more results.

Sites are read through this interface. To create, rename or delete a Site, ask your connected AI assistant (it shows a preview for your approval) or use the web or Android app. To retrieve Things under a Site, omit kind and filter by the chosen siteId.

Areas

Permission: inventory.read

An Area belongs to a Site, such as Kitchen in House. Look it up within its Site to avoid mixing identically named Areas from different Sites.

{
  "kind":"Area",
  "siteId":"11111111-1111-4111-8111-111111111111",
  "text":"Kitchen",
  "offset":0,
  "limit":100
}

Read the Area's returned id and siteId. To search Things in the Area, omit kind and supply its areaId; you may also supply its Site ID. If both filters are supplied, both must match. A mismatched pair returns no matching inventory rather than silently selecting another Area.

Areas are read through this interface. To create, rename or delete an Area, ask your connected AI assistant (it shows a preview for your approval) or use the web or Android app. An Area that still holds things cannot be deleted. In the web tester, changing Site refreshes the available Areas; an invalid old Area selection is cleared.

Categories

Permission: inventory.read

Categories classify Things independently of location. Find an existing Category before using its ID as a filter or assigning it in a supported create/edit batch.

{"kind":"Category","text":"Cameras","offset":0,"limit":100}

The lookup returns IDs for Categories already in the household. Resolve an ambiguous name rather than guessing. To search Things in that Category, omit kind and use categoryId.

This is a lookup interface. Category creation, schema/custom-field editing and deletion are not exposed. A Category lookup is not a complete export of all category settings or custom-field definitions.

Things, Parts and Contents

Permission: inventory.read

Search top-level Things

{
  "text":"camera",
  "categoryId":"33333333-3333-4333-8333-333333333333",
  "relationship":"Thing",
  "offset":0,
  "limit":100
}

The text search checks name, description, model, serial number and barcode using a case-insensitive substring match. It is not a semantic search and does not search attachment contents or notes. Combine filters only when every filter should apply.

Direct Contents of a Thing

{
  "parentThingId":"44444444-4444-4444-8444-444444444444",
  "relationship":"Content",
  "includeDescendants":false,
  "offset":0,
  "limit":100
}

For direct Parts, change relationship to Part. For all direct children, omit the relationship filter. Use includeDescendants:true to include nested descendants. A relationship filter still applies to each returned descendant. The parent itself is excluded.

relationship:"Thing" means top-level records only and excludes both Parts and Contents. Any, omitted or null includes all relationships. The model uses the singular Content in the request even though the user interface says Contents.

Important returned fields

This response is a defined projection, not the complete underlying Thing JSON. Notes, full custom fields, loan/rental/contact records and structured event histories are not returned by this query. Use the attachment endpoints to inspect supported attached content separately.

Valuations and missing values

Permission for reports: reports.read

Send the same relevant Thing filters with report:true. Omit kind.

{
  "areaId":"22222222-2222-4222-8222-222222222222",
  "relationship":"Thing",
  "report":true
}

A valuation report calculates across all matches on the server; you do not need to page through item lists to obtain its totals. A report's items is null, returnedCount is zero and hasMore is false, while totalCount still gives the matching-record count.

Illustrative response excerpt—not a full response:

{
  "totalCount":12,
  "returnedCount":0,
  "hasMore":false,
  "nextOffset":null,
  "summary":{
    "totalsByCurrency":{"GBP":1250.50},
    "missingValuations":2,
    "estimatedValuations":4,
    "enteredValuations":6
  },
  "items":null
}

Each effective value is multiplied by quantity. Currency totals remain separate. The counts describe records, not a sum of quantities.

An entered current value, including zero, takes precedence. Otherwise ThingzSafe may calculate an estimate from purchase price, purchase date, warranty/category rules and the household's new-for-old preference. Where the existing inventory rules permit it, undated Contents can have a purchase-price estimate; not every undated top-level Thing or Part has one. An estimate is not a fresh AI market valuation. This route does not call the Pro AI valuation service.

To list records with neither an entered nor an estimated value:

{"missingValueOnly":true,"offset":0,"limit":20}

By default, reports include all statuses and Parts/Contents. Specify relationship:"Thing" for top-level-only totals. The API does not apply all the web inventory valuation report's optional risk, exclusion or ownership filters, and it does not convert currencies.

Photos and documents

Permissions: inventory.read + files.read

List selected attachments

POST /v1/developer/attachments/list:

{
  "itemIds":["44444444-4444-4444-8444-444444444444"],
  "offset":0,
  "limit":50
}

Select 1–20 Thing IDs per request. The result is wrapped in isError, code, structuredContent and content. In structuredContent, inspect totalCount, returnedCount, hasMore, nextOffset and items.

Each file entry identifies its owning itemId, blobId, fileName, source, available mimeType and byte length, plus available. A file reference can exist even if its blob is unavailable. The same blob referenced repeatedly on one Thing is listed once; the same blob associated with two Things may appear for both.

Include the IDs of Parts/Contents separately when their attachments matter. A parent Thing alone does not implicitly include its children's files.

Read an attachment

POST /v1/developer/attachments/read:

{
  "itemId":"44444444-4444-4444-8444-444444444444",
  "blobId":"66666666-6666-4666-8666-666666666666",
  "page":1
}

Both IDs must come from the selected Thing and its file listing. The service checks the live relationship and household. It does not accept arbitrary storage URLs.

Illustrative response shape, with image data omitted:

{
  "isError":false,
  "code":null,
  "structuredContent":{
    "itemId":"44444444444444448444444444444444",
    "blobId":"66666666666646668666666666666666",
    "fileName":"Example receipt.pdf",
    "mimeType":"application/pdf",
    "pageCount":3,
    "firstPage":1,
    "returnedPages":2,
    "hasMore":true,
    "nextPage":3
  },
  "content":[
    {"type":"text","text":"Attachment metadata is supplied here."},
    {"type":"image","mimeType":"image/jpeg","data":"BASE64_JPEG_PAGE_1"},
    {"type":"image","mimeType":"image/jpeg","data":"BASE64_JPEG_PAGE_2"}
  ]
}

For this example, call again with page:3. PDF page numbering is one-based. Images and text documents use page 1 only.

Decode image blocks' base64 data as the stated MIME type. Text blocks contain extracted text. The returned images are prepared views, not an original-file download API. PDFs are rendered, so scanned documents can be inspected too. No public/SAS download URL is returned, and no AI analysis or receipt classification occurs on this API route.

Supported formats and omissions are listed in the ChatGPT files guide. In particular, DOCX is main-document text only, and XLSX/RTF are unsupported. Treat unavailable, unreadable, unsupported or partial reads as unexamined. Do not count them as negative evidence.

Create, edit and move Things

Use a two-stage flow: preview the exact proposal, show it to the user for approval, then execute the saved preview. Previewing does not change or lock the inventory while the user decides.

Each proposal needs a fresh requestId of 1–80 characters; a generated UUID is convenient. Use the same request ID and identical payload only for retries of that proposal. A developer batch is bound to the credential that created it. Use the same token for preview, execution and recovery.

Create top-level Things

POST /v1/developer/batches/preview:

{
  "requestId":"example-create-001",
  "action":"create",
  "siteId":"11111111-1111-4111-8111-111111111111",
  "areaId":"22222222-2222-4222-8222-222222222222",
  "categoryId":"33333333-3333-4333-8333-333333333333",
  "items":[{"name":"API test — delete later","quantity":1,"notes":"Created for API testing"}]
}

Use a real destination Site. The Area, if supplied, must belong to it. Omit Area to put the Thing directly in the Site. If Category is omitted, an existing Other category is used; if it is unavailable, choose an existing category explicitly. Creation is for top-level Things, not new Parts/Contents under a parent.

Edit existing Things

{
  "requestId":"example-edit-001",
  "action":"edit",
  "itemIds":["44444444-4444-4444-8444-444444444444"],
  "values":{"notes":"Checked through the API","needsReview":false}
}

The supplied values apply to every selected record. Omitted fields are preserved. An optional top-level categoryId can reassign the Category. Use a move action for location changes.

Supported values/create-item fields:

Name must be non-empty; ordinary text is limited to 200 characters and notes to 4,000. Quantity must be 0–1,000,000, with whole quantities for top-level Things and Parts; existing Contents may permit fractions. Money values must be non-negative and no greater than 1,000,000,000,000. Use ISO date strings such as 2026-10-25. Optional nullable fields can be cleared with null; quantity and currency cannot be null, and name cannot be empty/null.

This is not a general record-patching interface: do not send status, photo IDs, parent IDs, loan/disposal information or arbitrary custom fields. Existing unrelated data is preserved on supported edits.

Move Things and their owned children

{
  "requestId":"example-move-001",
  "action":"move",
  "itemIds":["44444444-4444-4444-8444-444444444444"],
  "siteId":"11111111-1111-4111-8111-111111111111",
  "areaId":"22222222-2222-4222-8222-222222222222"
}

A selected Thing moves with its owned Parts/Contents, including nested children. Inspect the full expanded preview: the 100-record batch limit includes affected descendants. Selecting both parent and child does not move the child twice. Moving a selected child makes it a top-level Thing at the destination; moving a parent retains the descendants' ownership links beneath it. This operation does not move stock into another Thing or split a quantity.

Review the response

A preview includes batchId, previewHash, status, expiresAt, action, affectedCount and changes. Each change includes an ID, expected revision, before and after details. A create has no before record.

Show all affected records and before/after details to the user. Resolve display names from the location/category lookups where needed. A preview expires after 15 minutes. Relevant changes to records or a moved ownership tree make execution fail; prepare a new preview and obtain approval again.

Execute the approved preview

POST /v1/developer/batches/execute:

{
  "batchId":"BATCH_ID_FROM_PREVIEW",
  "previewHash":"EXACT_HASH_FROM_PREVIEW",
  "confirm":true
}

The two placeholder strings must be replaced with the server's actual values. confirm:true is the calling application's assertion that approval was obtained; it is not independent proof of a human click. Your application must implement the review/approval step.

Execution is all-or-nothing. A successful result has status Executed, the action, affected count and record IDs/revisions. Changes join normal cloud sync and have integration Events. Devices will see them when they next sync.

Recover after a timeout

GET /v1/developer/batches/{batchId} with the same developer credential. If already Executed, use the saved result. Otherwise, retry the exact batch/hash while valid. Do not generate another create request merely because a response was lost. Re-executing a completed batch returns its original result. Old completed records may be cleaned up after 90 days; do not use them as indefinite job storage.

Limits, errors and unsupported operations

Current bounds

The shared minute allowance is used by inventory queries, attachment calls, new inventory previews and feedback preview/submit calls. Inventory execution and saved-batch recovery have their own checks; successful retries do not consume another committed-write allowance. Other HTTP rate limits also apply. Limits are not an entitlement to unlimited CPU, attachment analysis or ChatGPT usage. Treat explicit processing-limit errors as incomplete work rather than inventing totals.

Read failures correctly

Inventory query/write routes generally return a JSON code on application errors. Attachment routes use an isError/code wrapper as well as HTTP status. Inspect both the HTTP status and response body. MCP tools can return HTTP 200 with result.isError:true; inspect the tool result rather than assuming transport success means the operation succeeded.

Error names are route-specific; an error body need not contain a human-readable message. Log the code and request context, but keep tokens and attachment payloads out of application logs.

Not exposed in this release

Site/Area/Category creation or deletion; Thing deletion; creating new Parts/Contents under a parent; attachment uploads or original-file downloads; contact queries; structured loan/rental/reservation queries or actions; fitted/installation history; Pro AI recognition/valuation actions; arbitrary status/custom-field changes; currency conversion; public access to the support log. Use the ordinary app/web features where available.

MCP tools and advanced connection details

Most developers should use the developer-token routes above. MCP is the separate interface used by the ChatGPT, Claude and Google Gemini connections; it uses OAuth rather than a developer token. All three use the same server address. ChatGPT uses client ID thingzsafe-chatgpt and Claude uses thingzsafe-claude. Gemini connects through the published ThingzSafe app, so it needs no client details.

MCP endpoint: https://app.thingzsafe.com/mcp. It supports stateless Streamable HTTP POST with negotiated protocol versions 2025-03-26 and 2025-06-18. Tool discovery is public; private tool calls require authorisation.

Discovery endpoints are /.well-known/oauth-protected-resource, /.well-known/oauth-protected-resource/mcp and /.well-known/oauth-authorization-server. The static public client ID is thingzsafe-chatgpt; token endpoint authentication is none, with mandatory PKCE S256. The default configured callback is https://chatgpt.com/connector_platform_oauth_redirect. This is not a promise that arbitrary third-party OAuth client IDs or redirect URLs can register themselves.

Read the tool schemas rather than copying developer query bodies unchanged: for example, inventory_valuation selects report behavior through the tool name and does not accept a report field. list_inventory_locations uses kind; search_inventory selects Things without a kind argument. Follow isError, scopes and continuation fields.

Support reports map Bug to a failed supported operation and Change to a feature request. Both report tools require inventory.read plus feedback.write. Show the complete comment, destination and account/household association before submission. Submit the exact previewId/previewToken with confirm:true. The sealed token is connection-bound and expires in 15 minutes if unused. There is no equivalent developer-token feedback endpoint in this release.

Before you start

1. You are in a household with an active Household or Pro plan, as the owner or an invited member. Use your own ThingzSafe sign-in.
2. Your phone has finished its first sync. ChatGPT reads your cloud household, not records that exist only on a phone.
3. You have a ChatGPT account. It currently works with a free account, but OpenAI decides which plans can use plugins and custom connections, and may change this. If you cannot see the option, check your plan or ask your workspace administrator.
4. Use ChatGPT on the web, on a computer, to set the connection up.

You never paste a ThingzSafe password, a developer token or an OpenAI API key into the conversation.

Don’t have ThingzSafe yet? Get it on Google Play.

Choose the connection method

Use the manual setup below. It works fully and takes a few steps. Availability of custom connections is controlled by OpenAI and your workspace.

Connect manually using the server address

1. In ChatGPT, open the plugin or app management area and choose to add a custom MCP server. OpenAI currently calls this Plugins → + → Add custom MCP server. The labels may change.
2. Name it ThingzSafe.
3. Set the server URL to https://app.thingzsafe.com/mcp
4. Choose OAuth authentication. If it asks for client details, use client ID thingzsafe-chatgpt and leave the client secret empty.
5. Create the connection. When prompted, sign in to ThingzSafe, choose your household and read the permissions requested.
6. Tick any permission you want to allow (inventory changes, photos and documents, or reports), then allow the connection.
7. In a conversation, type @ and choose ThingzSafe, then ask a question about your inventory.

If you already have a ThingzSafe connection, refresh its tools to pick up new features. Reconnect when a new permission is requested.

Permissions

An existing connection does not gain new permissions automatically. File access does not imply inventory write access, and inventory write access does not imply permission to submit a support report.

Useful questions

• “What top-level Things are in the Kitchen? Exclude Parts and Contents.”
• “List the Contents of the fridge freezer in the Kitchen.”
• “Show the Parts recorded under my car.”
• “What are my workshop tools worth? Show estimated values and how many Things have no value.”
• “Find everything with ‘ladder’ in its name and tell me where each one is.”
• “Find my camera equipment, including Parts, and check its saved receipts.”

Give the Site and Area when a name could match several places. If more than one Thing or location matches, ChatGPT should ask you which one you mean.

Values and totals

An entered current value is used first. Where ThingzSafe can calculate a purchase-price estimate, that estimate counts as a value. A value of zero is also a known value. “No value” means neither an entered value nor an available estimate exists.

Totals multiply each value by quantity. Different currencies are kept separate; they are not converted into a single total. Tell ChatGPT whether you want top-level Things only or also Parts and Contents. The default search/report includes all statuses and child records, which may differ from a web report with additional filters selected.

Approve an inventory change

Ask, for example:

Preview changing the notes on “Garden ladder” to “Checked through ChatGPT”. Do not execute yet.

Check the exact Things and before/after details. If correct, tell ChatGPT to execute that preview. If not, request a corrected preview.

A preview holds no record locks and expires after 15 minutes. If the relevant inventory changes meanwhile, execution is rejected. Review a new preview before approving again. After a timeout, ask ChatGPT to check or retry the same batch; do not ask it to create the Thing again.

Disconnect

In the ThingzSafe web app, open How to use API, find the ChatGPT connection and choose Revoke. Revocation stops further authorised access. It does not undo inventory changes already committed, withdraw a submitted report or remove content already sent to a ChatGPT conversation.

Find a topic

Good to know

• ChatGPT, Claude and Google Gemini are separate services from OpenAI, Anthropic and Google. Their account access, connection availability and usage limits are set by those companies and your workspace.
• New to ThingzSafe? Get it on Google Play and set up your household first.
• You never give ThingzSafe an OpenAI API key, and you never paste a ThingzSafe token into a chat.
• ThingzSafe’s built-in AI photo recognition and replacement valuations are separate Pro features.

Ask for a specific check

Find my camera equipment, including Parts and Contents. Inspect its saved photos and documents. Tell me which Things have a matching receipt, which documents appear mismatched, and which Things remain uncertain or unexamined. Read every PDF page. Do not change anything.

Photo/document reading requires its own permission. ChatGPT first identifies the Things, lists their attached files, then reads the selected content. An attachment on a Part or Content item belongs to that child record; selecting only its parent does not automatically inspect the children's files.

What can be read?

Thing photos, document attachments, Event files and recognised embedded loan/disposal file references can be listed. This does not provide structured loan/rental queries.

Interpret the result

Ask ChatGPT to count Things with receipts, once per Thing, not the number of receipt files. A useful result separates:

• Confirmed receipt or proof-of-purchase matches.
• Purchase documents that appear to refer to a different item.
• No receipt found among all successfully inspected attachments, including Things with no saved attachments.
• Uncertain or unexamined Things, with the reason.

A filename containing “receipt” is a clue, not proof. A missing, unreadable, unsupported or oversized attachment is not evidence that no receipt exists. An unclear image may need you to open the original or provide a clearer copy. The attachment reader does not itself classify receipts or save its conclusions back to the inventory.

Large checks

ChatGPT may work through several batches. Lists and multi-page PDFs have continuation fields. If a limit is reached, it should explain which Things or pages remain and resume after the limit permits. It should not describe a partial inspection as complete.

File access sends the retrieved content to ChatGPT. Use the permission only for a household and files you want to analyse there.

If your question needs information that the connection does not provide, ChatGPT can explain the limitation and offer to draft a feature request. A supported operation that fails may instead warrant a bug report. No matching records in a search does not, by itself, mean a feature is missing.

1. Ask ChatGPT to prepare a report. Approve the separate report-submission permission if prompted.
2. Read the complete report text, whether it is a Bug or Change request, and the destination: the ThingzSafe bugs and requests log, visible to support administrators.
3. Ask for any corrections. Do not include passwords, access tokens or details you do not want support to receive.
4. If satisfied, approve submission. ChatGPT should only say it was submitted after the server returns a report ID.

Previewing or declining a report does not submit it. An unused preview expires after 15 minutes. The report is associated with your ThingzSafe account and household. A conversation transcript, reply email or attachment is not added automatically. You should knowingly approve any inventory or other personal details included in the actual text.

After a timeout, retry the same preview so the original report ID can be recovered without a duplicate. Each household can submit up to ten ChatGPT reports in a rolling 24 hours. This feature submits feedback; it does not provide access to other users' reports or add the requested inventory capability immediately.

ChatGPT says it cannot do something, but does not offer to send us a request.

Refresh the ThingzSafe connection's tools in ChatGPT, then start a new conversation with ThingzSafe selected. You can also ask directly: “Prepare a feature request for this. Do not submit it yet.” A report is something you ask ChatGPT for in the conversation. There is no separate button for it in ThingzSafe.

If report permission is missing, reconnect and explicitly approve Submit bug reports and feature requests. Inventory write permission alone does not grant report submission. After reconnecting, request a new draft, review the complete text and destination, and approve only if correct. Do not recreate an already submitted report just to test a retry; use its original preview.

A newly added tool is missing.

Refresh the plugin's tool list. Reconnecting alone and refreshing tool definitions are different operations. Start a new conversation if the previous one still appears to use old capabilities. Your ChatGPT account or workspace may also restrict custom connections or individual tools.

The API tester works but ChatGPT cannot use the connection.

The tester uses a developer token; ChatGPT uses its own OAuth connection. Check that ThingzSafe is selected in the conversation, the correct household was authorised and the connection has the necessary scopes. Do not paste the developer token into chat.

Photos or documents are unexamined.

Check file permission, the exact owning Thing, supported format, page continuation and limits. Open the original in ThingzSafe if the prepared view is unclear. Unexamined does not mean no receipt.

A preview or connection changed before approval.

Ask for a new preview and review it again. For an uncertain response after execution/submission, recover or retry the same original batch/report first, to avoid duplication.

Before you start

• Install ThingzSafe from Google Play on your Android phone or tablet.
• You need an active ThingzSafe Household or Pro plan, or to have joined someone’s household.
• Set up your household in the Android app and let it finish syncing. Your assistant works with the cloud inventory, not records that exist only on a phone.
• Have your usual ThingzSafe sign-in email address ready. Use the same address as in the app.
• Use a web browser on a computer for the first connection. You need a Claude account.

ThingzSafe does not include a paid subscription for any assistant. Any account requirements or charges from your chosen assistant are separate.

Connect Claude

1. Open claude.ai and sign in.
2. Open Customize in the left menu, then Connectors.
3. Choose + Add, then Add custom connector. If you see Custom → Web, choose that.
4. Name it ThingzSafe and enter this server address: https://app.thingzsafe.com/mcp
5. Continue to the sign-in settings. Choose Sign in now and Use your own OAuth client. If you see an Advanced settings section, open it to find the client fields.
6. Enter thingzsafe-claude in Client ID. Leave Client secret blank. The client ID is a connection label, not a password.
7. Finish adding the connector, then choose Connect if prompted.
8. On the ThingzSafe page, sign in if needed. Check the heading says Connect Claude, choose your household and review the permissions.
9. Approve the connection to return to Claude.

If you use Claude through a work organisation, its administrator may need to add the connector before you can connect your own account. Claude’s menus and button names change from time to time, so follow the closest match if yours look slightly different.

Use the connection in a conversation

Start a new conversation. If ThingzSafe is not already switched on, open the + menu beside the message box, choose Connectors and turn on ThingzSafe. Then ask your first question.

Choose what the assistant can do

On the ThingzSafe permission screen, choose the household you want to connect, then read the permissions before you allow access.

• Read inventory: find your Things and their details.
• Read valuation reports: answer questions about the values already recorded.
• Make changes: prepare and apply changes you have approved.
• Read photos and documents: inspect supported files already saved in ThingzSafe.
• Send bug reports or suggestions: submit a report only after you have checked and approved its contents.

You may see only some of these. Changes, file access and sending reports each have their own permission. A read-only connection cannot make changes. If a later request needs more permission, you may be asked to connect again and approve it.

You sign in on ThingzSafe’s own page. Never type your ThingzSafe sign-in codes, passwords or access tokens into the conversation.

Try your first question

Start with something simple:

• “List my sites in ThingzSafe.”
• “Where is my cordless drill?”
• “Show my reminders for the next 20 days.”
• “Show the notes for the goat shed.”

If you have similar names in several places, include the location, for example “Find the drill in the workshop at Home.” If more than one thing matches, the assistant should ask which one you mean.

Check changes before saving

You can also ask for a change, such as “Add a note to the drill saying the battery was replaced today.”

1. The assistant prepares a preview.
2. Check the record, location and proposed details.
3. If they are right, approve it. If not, ask for a correction first.

Nothing in your inventory changes just because a preview has been prepared. If a save seems to take a long time, ask the assistant to check whether the change completed before asking it to create the same thing again.

Work with your tables

Ask the assistant to explain or compare the information in your ThingzSafe tables, for example “Compare the number of goats with the feed bought each month.” Include the table name and location if needed.

With Pro, you can ask it to create a table from a spreadsheet, or from a table you paste into the conversation. Tell it which Site, Area or Thing the table belongs to, and check the column names, dates, numbers and preview before approving. Larger tables may need several smaller batches, each with your approval. Spreadsheet formatting and working formulas are not carried across.

Photos, receipts and documents

With the photos and documents permission, the assistant can inspect supported files already saved in ThingzSafe. Ask it to find the item and read the relevant document. A large document may need to be read in sections. See Check receipts, photos and documents for what can be read.

Adding a new attachment through this connection is not supported at present. Save it with the ThingzSafe Android or web app first, or use your household’s email receipts address.

What the connection can access

The connection uses the household you selected and the permissions you approved. It follows your ThingzSafe plan, so Pro features still need Pro. It never gives access to another household’s inventory.

Information returned to the assistant is handled under that provider’s account settings and privacy terms, as well as ThingzSafe’s. Choose what you ask it to read, particularly for personal documents. Claude can make mistakes, so check its answers and any proposed changes.

Disconnect from ThingzSafe

Sign in to the ThingzSafe web app, open How to use API and find AI connections. Choose Revoke beside the connection you want to stop.

Revoking stops its access to ThingzSafe. It does not delete your inventory or disconnect your other assistants, and it does not remove information already returned in earlier conversations.

If it does not connect

• No Connectors option: check the Customize menu. On a work account, ask your administrator whether custom connectors are allowed.
• Asked for a secret: check that you chose your own OAuth client, entered thingzsafe-claude and left the secret blank. Never use your ThingzSafe password here.
• No household to choose: finish setting up or joining your household in the Android app, let it sync, and check you used the same email address.
• Empty or old results: check syncing has finished and the correct household is connected.

If it still fails, contact us with the message or error reference. Never send sign-in codes, passwords or tokens.

Related help

Connect Google Gemini · Connect ChatGPT · Tables help · All AI assistants and the API

Before you start

• You need an active Household or Pro household.
• The household owner creates the receiving address, once, and manages which email addresses are allowed to send to it.
• Emails are accepted from registered email addresses: the verified sign-in addresses of people the owner has admitted to the household. An invitation that has not been accepted is not enough.
• Orders can be in the body of the email, or come with a PDF or photo of the receipt attached.

Step 1. Create your household’s receiving address

The household owner does this once.

1. Sign in at app.thingzsafe.com and check you are in the right household.
2. Open Email receipts in the menu on the left.
3. Choose Create receiving address.
4. Your household’s full address appears on the screen. Choose Copy address to copy it.

This address belongs to your household. It is not the email address you sign in with. Copy all of it, including everything after the @. You do not need to set up another email account or a separate mailbox. Always use the complete address shown in ThingzSafe.

Who can send to it?

Any registered email address can forward to it. That means the email address someone uses to sign in to ThingzSafe, once they have joined your household and verified it. The household owner admits and manages these addresses in People and devices. Emails from any other address are not accepted.

Step 2. Forward an order or receipt

1. Open your usual email app, using a registered email address (your ThingzSafe sign-in address).
2. Forward the order or receipt to the address you copied. Keep any attached PDF or photo of the receipt.
3. Send it once, then wait a short while for it to be processed.
4. Go back to Email receipts and choose Refresh. Then open your Review queue to see any suggestions.

What works best

• A clear PDF, or a clear, well-lit photo of the receipt.
• An order whose details are written in the email itself.
• The real file attached. A link to an online receipt does not work, because ThingzSafe does not open links to fetch purchases.

Size limits

• The whole email should be under 8 MB, with no more than 10 attachments. Email encoding makes files bigger, so keep them comfortably below that.
• Only the first five pages of a PDF are read.
• Normal file-safety checks and your household’s storage limits also apply. Some files can be rejected even if your email provider lets you send them.

Step 3. Check the suggestions and save

1. Open Review queue and choose a suggestion.
2. Compare its name, supplier, purchase date, quantity, currency and purchase price with the original. Correct anything that is missing or wrong, and read the notes.
3. Choose the Site, Area and Category the screen asks for.
4. Check the item is not already in your inventory. If the same purchase is forwarded in two emails, you will get two suggestions.
5. Save only the items you want. Then open the saved Thing and check its details and the kept files under Photos & files.

What Household and Pro do

• Household gives you basic text reading. It reads the text in the email and in PDF or image attachments, using OCR (which reads the words in a picture) for images. It does not use AI to interpret the email.
• Pro does the same, then adds AI to interpret the email text and the attachments it accepts. It can suggest several products from one email.

With either plan, the original email and accepted attachments are kept privately as source material, and nothing is created or merged without you. Check all the details. Unknown or unclear details may be left blank or put in the notes for you to sort out.

The Email receipts screen

The screen explains how it works at the top, then shows your receiving address in full. The address buttons sit beside the address. Results from processing appear below.

If something goes wrong

There is no receiving address

Check you are signed in as the household owner and that your plan includes Email receipts, then choose Create receiving address. If the feature is not available, it may not be switched on for your household yet.

I sent an email but no suggestion appears

Choose Refresh and read the processing result. Check you used the current address, and sent from a registered email address. Check the attachment size, type and number of PDF pages. If processing finished, look in the Review queue before you send again.

My email or address was rejected

Addresses that are not registered with the household, invitations not yet accepted and unverified email addresses cannot send receipts. A reset address also stops accepting emails. Use the current address and check People and devices. Your email provider accepting the message does not mean ThingzSafe accepted it.

Some attachments were rejected

Read the warnings. Other attachments in the same email may still be processed, so do not assume every file was read or kept. Try a clear PDF or image within the limits.

A detail or amount is wrong

Correct it in the review. Small print, blurry photos, handwriting and unusual layouts can reduce accuracy. Reading text from images currently expects English, whatever language the app is set to. Pro can interpret readable foreign-language text, but that does not guarantee accurate results from every language or image.

A message stays failed or unclear

Note the reference shown, the time you sent it and the status reported, then contact us. Please do not keep forwarding or retrying an unclear message, and never include passwords or sign-in codes.

Related help

Review queue on the web · People, invitations and devices · Photos and files on the web · Plans, purchases and cancellation

Send a useful report

Open Report a bug or suggest a change in Android or the web app. Choose the report type, describe what happened and include steps to reproduce it. Say what you expected and what actually happened.

You can attach one image, up to 4 MB. Choose an image file and check it contains no private details. A screenshot is often more useful than a long description. Remove it or replace it before sending if needed.

Include an email address if you want a reply, where the form offers it. When available, the report includes app/platform information and your household ID so support can trace the relevant backend records. A household ID is not required for users without a household.

Errors and support references

Read the plain-language message and follow its suggested next step. Use Details or Copy details if offered, and include the support reference with your report. You do not need to interpret HTTP codes or internal processing stages.

Do not send passwords, sign-in codes or payment details. Do not include household data unrelated to the problem.

View the development board

The website development board shows published changes and requests. A report being logged does not guarantee a release date or that every suggestion will be implemented.

Before you start

• Install ThingzSafe from Google Play on your Android phone or tablet.
• You need an active ThingzSafe Household or Pro plan, or to have joined someone’s household.
• Set up your household in the Android app and let it finish syncing. Your assistant works with the cloud inventory, not records that exist only on a phone.
• Have your usual ThingzSafe sign-in email address ready. Use the same address as in the app.
• Use a web browser on a computer for the first connection. You need a Google account.

ThingzSafe does not include a paid subscription for any assistant. Any account requirements or charges from your chosen assistant are separate.

Connect Google Gemini

1. Open gemini.google.com and sign in to your Google account.
2. Open the connected-apps list. Depending on your Gemini layout, this is under Customise → Connectors or Settings & help → Connected apps.
3. Find ThingzSafe in the available apps and choose it.
4. Choose Connect and follow the sign-in prompts.
5. On the ThingzSafe page, sign in with your usual ThingzSafe email address. Check the heading says Connect Gemini.
6. Choose your household, read the permissions and approve the connection to return to Gemini.

You do not need to create a custom connection, enter a client ID or obtain a client secret. Just use the ThingzSafe entry in the available apps. Gemini’s menus and button names change from time to time, so follow the closest match if yours look slightly different.

Use the connection in a conversation

Start a new conversation and type @ThingzSafe. Choose ThingzSafe when it appears, then ask your question, for example “@ThingzSafe list my sites.”

If ThingzSafe does not appear, check the troubleshooting list below.

Choose what the assistant can do

On the ThingzSafe permission screen, choose the household you want to connect, then read the permissions before you allow access.

• Read inventory: find your Things and their details.
• Read valuation reports: answer questions about the values already recorded.
• Make changes: prepare and apply changes you have approved.
• Read photos and documents: inspect supported files already saved in ThingzSafe.
• Send bug reports or suggestions: submit a report only after you have checked and approved its contents.

You may see only some of these. Changes, file access and sending reports each have their own permission. A read-only connection cannot make changes. If a later request needs more permission, you may be asked to connect again and approve it.

You sign in on ThingzSafe’s own page. Never type your ThingzSafe sign-in codes, passwords or access tokens into the conversation.

Try your first question

Start with something simple:

• “List my sites in ThingzSafe.”
• “Where is my cordless drill?”
• “Show my reminders for the next 20 days.”
• “Show the notes for the goat shed.”

If you have similar names in several places, include the location, for example “Find the drill in the workshop at Home.” If more than one thing matches, the assistant should ask which one you mean.

Check changes before saving

You can also ask for a change, such as “Add a note to the drill saying the battery was replaced today.”

1. The assistant prepares a preview.
2. Check the record, location and proposed details.
3. If they are right, approve it. If not, ask for a correction first.

Nothing in your inventory changes just because a preview has been prepared. If a save seems to take a long time, ask the assistant to check whether the change completed before asking it to create the same thing again.

Work with your tables

Ask the assistant to explain or compare the information in your ThingzSafe tables, for example “Compare the number of goats with the feed bought each month.” Include the table name and location if needed.

With Pro, you can ask it to create a table from a spreadsheet, or from a table you paste into the conversation. Tell it which Site, Area or Thing the table belongs to, and check the column names, dates, numbers and preview before approving. Larger tables may need several smaller batches, each with your approval. Spreadsheet formatting and working formulas are not carried across.

Photos, receipts and documents

With the photos and documents permission, the assistant can inspect supported files already saved in ThingzSafe. Ask it to find the item and read the relevant document. A large document may need to be read in sections. See Check receipts, photos and documents for what can be read.

Adding a new attachment through this connection is not supported at present. Save it with the ThingzSafe Android or web app first, or use your household’s email receipts address.

What the connection can access

The connection uses the household you selected and the permissions you approved. It follows your ThingzSafe plan, so Pro features still need Pro. It never gives access to another household’s inventory.

Information returned to the assistant is handled under that provider’s account settings and privacy terms, as well as ThingzSafe’s. Choose what you ask it to read, particularly for personal documents. Gemini can make mistakes, so check its answers and any proposed changes.

Disconnect from ThingzSafe

Sign in to the ThingzSafe web app, open How to use API and find AI connections. Choose Revoke beside the connection you want to stop.

Revoking stops its access to ThingzSafe. It does not delete your inventory or disconnect your other assistants, and it does not remove information already returned in earlier conversations.

If it does not connect

• Asked for a client ID or secret: you have opened the custom-connection form. Go back and find the ThingzSafe app in the available apps instead.
• ThingzSafe is missing: check the connected-apps list, then contact us if it is still unavailable.
• Access has expired: use the connection’s sign-in or reconnect option and review the ThingzSafe permissions again.
• No household to choose: finish setting up or joining your household in the Android app, let it sync, and check you used the same email address.
• Empty or old results: check syncing has finished and the correct household is connected.

If it still fails, contact us with the message or error reference. Never send sign-in codes, passwords or tokens.

Related help

Connect Claude · Connect ChatGPT · Tables help · All AI assistants and the API

Sign up when you want to hear from us

No spam, sales pitches or marketing. The list is for ThingzSafe feature news, release updates and how-to articles. You can unsubscribe using the link in any update.

On Android, open Main menu → App and choose the news signup. The web app and website offer the same signup link. Joining the list is optional and separate from your ThingzSafe account or plan.

Sign up for ThingzSafe updates

Sender hosts the signup form and manages the email list. Enter the email address you want updates sent to. Complete any confirmation email the form sends.

What each plan includes

Free: unlimited local Android inventory with locations, Things, Parts, Contents, photos, reminders, labels and backups. No cloud account is needed.

Household: all Free features plus household sync, sharing with invited people, the web app, bulk editing and imports, email receipts with basic text reading, connections for ChatGPT, Claude and Google Gemini, and developer API access. Only the household owner pays; invited members do not need a separate subscription.

Pro: all Household features plus category customisation, sale/disposal records, loans, rentals and reservations, Alexa voice control, Android bulk AI recognition and AI replacement valuations on Android and web, AI reading of emailed receipts and data tables with scheduled readings.

See the complete version comparison and prices.

Buy and activate

In Android, open Menu and the Account section to view plans and household setup. Google Play shows the current price, billing period, local taxes and renewal terms before purchase.

Use the ThingzSafe identity you intend to keep. After purchase, leave the app open while household setup and the first sync finish. If interrupted, use the setup retry or Restore purchases rather than buying again.

The web app is for an already provisioned Household or Pro account. Create or join the household through Android first.

Join an invited household

You do not need to buy your own subscription. Install ThingzSafe, then use household setup to join with the invited email address and invitation code. Back up any local inventory before joining or restoring, because a replacement operation can replace the inventory on that phone.

Restore, manage or cancel

Restore purchases checks the Google Play account that made the purchase. Use the same ThingzSafe sign-in identity for the household. On a new phone, use the cloud restore workflow carefully.

Manage renewal or cancel through Google Play subscriptions. Cancellation stops future renewal; current access follows the subscription terms. Local Free tools remain available. Keep an encrypted backup or export before paid access ends.

For sign-in or purchase problems, contact support with the reference shown. Never send passwords, sign-in codes or payment-card details.

A quick example

You keep chickens. You make a table called Monthly chicken records on your Barn Thing, with three columns: Month (a date), Chickens (a number) and Feed bought (a cost). Each month a new row appears, with the month already filled in. You add the figures and save the row, and you have a running record you can sort, export or ask ChatGPT, Claude or Gemini about.

Open Tables

• On Android: open the Site, Area or Thing, then choose Tables · Pro. You will find it after Notes.
• On the web: open the record and choose the Tables tab, after Notes. For Sites and Areas the tabs are under the title.

For help while you are working, use the usual Help button at the top of the app.

Create your first table

1. Choose Add table. The designer opens with one column and one empty row.
2. Type a Table name, such as Monthly chicken records.
3. Select the Column 1 ▾ heading. In Column settings, give the column a name and choose its type. Number columns offer an optional unit and a choice of what the number means. A Choice column asks for its list of choices. Choose Save to go back to the grid.
4. Use the Insert column button beside the column headings to add another column, and the Insert row button beneath the grid to add another row.
5. If you like, type your first readings straight into the grid.
6. Choose the main Save when the table is ready. Until you do, your columns, schedule and draft rows stay in the designer and are not saved.

Column settings also have buttons to move a column left or right, or Delete it. A table must keep at least one column. Removing a column asks you to confirm, and takes effect when you save the layout.

To change a table later, choose Table Layout to change columns. Within the layout editor, select a column heading to change that column. Saved readings show as a preview while you edit the layout. To change the readings themselves, save the layout first, then edit the rows.

Choose the right column type

What does your number mean?

For a number column, say what the figures represent. This helps reports read them correctly.

• Current count records how many you have right now; earlier entries stay in your history. This is a snapshot, such as the number of chickens today.
• Amount to add up is something you add up over a period, such as feed bought or money spent.
• Meter reading is a running total, such as tractor hours.

A blank cell means nothing has been entered. That is different from zero.

Start from another table

If another Thing, Site or Area already has the layout you want, you can copy it.

1. When adding a table, choose Based on another table….
2. Search for the record that has the layout, select it, then select its table. On the web, choose Save in that window to load it into the designer.
3. Check the copied grid, then choose the main Save to create your new table.

This copies the columns into a new table of its own. It does not copy any readings or reminders, and its schedule starts as Manual. If you have already entered an unsaved layout or readings, you are asked before they are replaced.

Have dated rows added for you

For a record you keep regularly, ThingzSafe can add a new row on each due date and remind you to fill it in.

1. Add a column and choose Date as its type.
2. Select the Auto Schedule button above the grid to open Recording schedule.
3. Choose how often, and which date column to fill. Enter or choose the first date, and check the time zone.
4. Choose Save to go back to the designer, then the main Save to apply your layout and schedule.

The first date must fit the schedule you chose. Only one date column can drive a table’s schedule. Any other date columns stay manual.

How the rows appear

• Our cloud service creates each row as its date becomes due, normally on its next background check. Your phone must sync before the new row shows there.
• If dates were missed, it catches up in batches of up to 31 dates per table. It sets one grouped reminder per table, not one for every missed reading.
• Changing a schedule keeps your earlier rows. Choosing Manual stops new automatic rows, but any readings already waiting still need completing.
• If you delete a scheduled row, it is not created again.

Fill in a scheduled row

1. Open the reminder. It takes you to Tables.
2. Choose the table, then Edit row on the row that is due.
3. Enter a value and choose Save. There is no separate completion checkbox.

Once your devices sync, a completed row no longer counts as pending. The reminder closes when every scheduled reading is complete. Saving any value clears that row’s notification. Zero and an explicitly selected No count as values; the automatically filled schedule date does not. Blank rows stay pending.

Read and edit a table

• Sort: select a column heading to sort. Select it again to reverse the order. To edit columns, choose Table Layout (Pro).
• See more: scroll sideways for more columns. On the web the grid scrolls on its own and keeps the headings in view. Use Next rows for more pages. Android also has Previous rows.
• Add: the Insert row button beneath the grid adds a row. To insert a column, first choose Table Layout.
• Change a reading: choose Edit row.
• Delete: Delete row removes one reading. Delete table removes the whole table and all its readings on every device. Delete a table before you delete the Site, Area or Thing that owns it.

Each row syncs separately. If another device has changed a row or its columns before you save, use the usual conflict and reopen steps instead of overwriting their work. If a save fails, your draft stays open so you can try again.

If you leave the designer without saving, the web asks whether to discard your draft. On Android, Cancel asks you to confirm only if there are changes to discard, and the normal Back action leaves without saving. Neither keeps a draft to resume later.

You can only change a column’s type if its existing values still fit. Removing a column that has data asks you to confirm, so export first if you want a copy.

Ask an AI assistant about your readings

Once ChatGPT, Claude or Google Gemini is connected to ThingzSafe, you can ask things like:

• “Compare chicken numbers with feed purchased each month.”
• “Add one to this month’s chick count.”
• “Add a row for tomorrow with the lamb count at seven.”

Say which Thing and table you mean if more than one could match. Your assistant shows the change it plans to make and asks for your approval first. “Add one” is different from “set the count to one”, and an empty number must be filled in before it can be increased. Tables do not add any new Alexa voice commands.

Export and reporting

Choose Export CSV to save or share the table for a spreadsheet. The column headings include the units you entered. An authorised reporting API also provides table definitions and rows, a page at a time. A ready-made Power BI connector is not included.

Plans and your data

• Creating tables, editing them and scheduled rows all need Pro.
• If you move down to an active Household plan, your existing tables can still be read and exported.
• If all paid cloud access ends, the normal cloud-access rules apply. Data already downloaded to Android stays on the device.
• For now, a household can have up to 50 columns in a table, 100 tables and 19,000 rows in total. The usual record-size and sync limits also apply.

If something looks wrong

A scheduled row has not appeared

Rows are created by our cloud service on its next background check, and your phone has to sync before it shows them. Check your connection and choose to sync, then look again.

My changes were not kept

Changes to columns, the schedule and draft rows are only saved when you choose the main Save. If the save failed, the draft stays open so you can try again.

I cannot create or edit a table

Creating and editing need a Pro plan. On a Household plan you can still read and export existing tables.

A column will not change type or be removed

The new type must suit the values already in the column. Removing a column that holds data needs confirming. Export first if you want to keep a copy.

Related help

Calendar and repeating reminders · Connect and use ChatGPT · Plans, purchases and cancellation · What Data tables can do

Layout and sorting

Choose Table Layout to change columns, or Auto Schedule to set when new dated rows appear. Insert-column and insert-row buttons use a small table with a right or down arrow. Insert column is available only in the layout editor. When viewing your data, select a column heading to sort; select it again to reverse the order. Saving an edited row keeps its table open.

Unit (optional) tells you what a number measures, such as kg, litres, hours or miles. Leave it blank if the column name makes that clear.

Find a table quickly

Search for a table’s name in the app. For example, search for goats to find your Goats table. Search for table or tables to see all tables you can access. Choose a result to open it on the Thing, Site or Area it belongs to.

The first table opens when you choose the Tables tab. Select another table by name. Click a column heading to sort; an arrow appears only on the column currently used for sorting.

Select the rows

1. Open Bulk Review Items and tick the AI valuation column for the rows you want. The header checkbox selects or clears all rows.
2. Click Generate AI valuations for selected items · Pro. Save pending grid edits when asked, so renamed items use their updated details.
3. Check the country and currency in the valuation window, then Generate. Keep the window open while it processes.

Review and apply

Select a low, middle or high comparable using its radio button, and open the VIEW link to check the listing. The proposed acceptance value updates with your choice. Edit and accept is available when you want a different reviewed amount.

Tick Mark item as reviewed when accepted if you want its Needs review flag cleared too. Accept individual proposals, or use Accept selected reviewed proposals to apply the selected results and close the window. Accepted values return to the current-value fields in the grid.

After generation, acceptance becomes the main action. A second accept action below the stack saves scrolling. Accepted item panels fold down.

Read the full guide to valuation basis, confidence, country, delivery and recovery.