Taggun

Delete campaign validation settings via DELETE /api/validation/v1/campaign/settings/delete/{campaignId}. DESTRUCTIVE AND IRREVERSIBLE: delete one campaign's validation settings entirely. The merchant allow and block lists, product codes, date window, balance range and fraud switches it held are gone, every later validation of that campaign loses the rules it was judged against, and Taggun documents no recycle bin, no restore and no soft-delete state. There is also no sandbox in which to try it first. PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.

Export known merchants via GET /api/account/v1/known-merchants/export. Export the account's known-merchant list -- the merchant names Taggun matches a receipt against when it normalises the merchant it read. Reads the account's own configuration; no receipt is processed.

Export known product codes via GET /api/account/v1/known-product-codes/export. Export the account's known-product-code list -- the codes Taggun matches line items against when enriching a receipt. Reads the account's own configuration; no receipt is processed. PLAN-GATED: product and category enrichment requires a Startup plan or above. A refusal here is an entitlement the account has not bought, not a rejected key.

Export product categories via GET /api/account/v1/product-categories/export. Export the account's product-category list -- the categories Taggun assigns line items to when enriching a receipt. Reads the account's own configuration; no receipt is processed. PLAN-GATED: product and category enrichment requires a Startup plan or above. A refusal here is an entitlement the account has not bought, not a rejected key.

Get campaign validation settings via GET /api/validation/v1/campaign/settings/{campaignId}. Read one campaign's validation settings by campaign id: its merchant allow and block lists, product codes, product line-item rules, date window, balance-owing range, smart-validate options and fraud-detection switches (digital detection, tamper detection and similarity checking). PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.

List campaign validation settings via GET /api/validation/v1/campaign/settings/list. List every campaign the account has validation settings for. A campaign holds the rules a submitted receipt is judged against -- allowed and blocked merchant names, product codes, a date window, a balance-owing range, and which fraud checks run. PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.

Submit extraction feedback via POST /api/account/v1/feedback. Correct what Taggun read off a receipt it has already processed, naming it by the `referenceId` the original request was tagged with, and supply the right merchant name, date, total, tax, currency or master category. This is training feedback for the account's own model, not a re-read of the receipt.

Add a known merchant name via POST /api/account/v1/merchantname/add. Add one merchant name to the account's known-merchant list, so later receipts from that merchant normalise to it. Changes the account's configuration for every future extraction, not just the next one.

Read a receipt from a base64 image (simple) via POST /api/receipt/v1/simple/encoded. Run OCR over a receipt or invoice supplied as a BASE64-ENCODED image in the request body (`image` with its `filename` and `contentType`) and return the simple result: the merchant, date, total, tax and currency Taggun read, without the per-field confidence and region detail the verbose form carries. This is the bounded way to send a file to Taggun through this integration -- the multipart `/simple/file` endpoint is not shipped. SPENDS A SCAN against the Taggun plan's allowance, against PRODUCTION -- Taggun publishes no sandbox environment, so the same key that reads also processes real receipts and is metered for them.

Read a receipt from a URL (simple) via POST /api/receipt/v1/simple/url. Run OCR over a receipt or invoice TAGGUN FETCHES ITSELF from an https URL you supply, and return the simple result. Taggun dereferences the URL, not Agentic Fabriq, and only https is accepted; the file must be reachable by Taggun's own servers. SPENDS A SCAN against the Taggun plan's allowance, against PRODUCTION -- Taggun publishes no sandbox environment, so the same key that reads also processes real receipts and is metered for them.

Read a receipt from a base64 image (verbose) via POST /api/receipt/v1/verbose/encoded. Run OCR over a receipt or invoice supplied as a BASE64-ENCODED image in the request body and return the verbose result: every extracted field with its confidence, the text region it came from, the full transcribed text, and -- with `extractLineItems` -- the product line items. This is the bounded way to send a file to Taggun through this integration; the multipart `/verbose/file` endpoint is not shipped. SPENDS A SCAN against the Taggun plan's allowance, against PRODUCTION -- Taggun publishes no sandbox environment, so the same key that reads also processes real receipts and is metered for them.

Read a receipt from a URL (verbose) via POST /api/receipt/v1/verbose/url. Run OCR over a receipt or invoice TAGGUN FETCHES ITSELF from an https URL you supply, and return the verbose result with per-field confidence, text regions and optional line items. Taggun dereferences the URL, not Agentic Fabriq, and only https is accepted. SPENDS A SCAN against the Taggun plan's allowance, against PRODUCTION -- Taggun publishes no sandbox environment, so the same key that reads also processes real receipts and is metered for them.

Validate a campaign receipt from a URL via POST /api/validation/v1/campaign/receipt-validation/url. Fetch a receipt from an https URL and judge it against a STORED campaign's settings, named by `campaignId`, rather than against rules passed in the call. The campaign's own allow and block lists, date window, balance range and fraud switches decide the verdict, so changing the campaign changes what this tool accepts. SPENDS A SCAN against the Taggun plan's allowance, against PRODUCTION -- Taggun publishes no sandbox environment, so the same key that reads also processes real receipts and is metered for them. PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.

Create campaign validation settings via POST /api/validation/v1/campaign/settings/create/{campaignId}. Create the validation settings for one campaign id: merchant allow and block lists, product codes, product line-item rules, a date window, a balance-owing range, smart-validate prompts and the fraud-detection switches. These rules govern every later validation of that campaign, so this changes how receipts submitted afterwards are judged. PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.

Validate a receipt from a URL via POST /api/validation/v1/receipt-validation/url. Fetch a receipt from an https URL and judge it against validation rules supplied IN THE REQUEST (`validation`) rather than against a stored campaign: merchant allow and block lists, product codes and line items, a date window, a balance-owing range, and the fraud checks -- digital detection, tamper detection and similarity checking. Returns the extraction and the per-rule verdicts. SPENDS A SCAN against the Taggun plan's allowance, against PRODUCTION -- Taggun publishes no sandbox environment, so the same key that reads also processes real receipts and is metered for them. PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.

Replace campaign validation settings via PUT /api/validation/v1/campaign/settings/update/{campaignId}. REPLACE one campaign's validation settings. This is a PUT and it takes the whole settings object, so a field left out of the body is a field removed from the campaign, not a field left alone -- read the campaign first and send it back changed rather than sending a partial update. PLAN-GATED: receipt validation and campaigns require an Advanced plan or above with Campaign Validation enabled in the account's Feature Settings. A refusal here is an entitlement the account has not bought, not a rejected key.