Dolibarr Module v1.2.4 Dolibarr 21 EN · FR · DE · ES · IT

Data Trust & Governance

36 documented data-quality checks across your Dolibarr records, a review queue to work every issue, and a fix that only counts when a later scan proves it. Read-only. No AI. No telemetry.

36
documented checks
9
record areas
5
quality dimensions
0
records changed

Watch the full demo

Data Trust & Governance for Dolibarr demo video Watch on YouTube ↗

Policy presets, an assessment, the issues queue, an issue with its evidence, duplicates, exceptions and the management report.

Why businesses need it

Bad data costs money long before anyone notices

Duplicate customers, broken links and wrong bank details don't show up on a dashboard. They show up as a wrong invoice, a payment to the wrong account or a report nobody believes.

Duplicates and incomplete master data

The same customer created twice, products sharing a reference or barcode, missing payment terms and codes. Every downstream document inherits the error.

Broken records and payment risk

Invoices pointing to third parties that no longer exist, net + tax that doesn't equal gross, invalid IBANs and one bank account shared by several suppliers.

Clean-ups that don't stick

A one-off clean-up spreadsheet has no owner, no proof and no follow-up. Six months later the same problems are back and nobody can say what was fixed.

How it works

Choose, assess, review, fix, verify

The module never repairs data itself. People fix records in Dolibarr, and the next scan proves the fix.

1

Choose the policy

Pick a business-type preset (distribution, retail, services, manufacturing, small business) and adjust any check.

2

Run an assessment

Whole-population scans in batches, within time and memory budgets. Nothing is written to ERP tables.

3

Publish issues

Each failing record opens an issue, updates an existing one, or is flagged as “came back” after a fix. Evidence is stored with a checksum.

4

Work the queue

Assign owners, review, snooze, dismiss with a reason, request an exception, and open the record to fix it in Dolibarr.

5

Verified fixed

An issue closes only when a later complete scan shows the same record passing the same rule. Nobody can simply mark it fixed.

Features

Governance your auditors can follow

Owners, evidence, exceptions and history, all inside Dolibarr. CSV export everywhere.

Dashboard

Setup checklist, KPIs (active, high priority, coverage, without owner, verified fixed), trend of flagged vs evaluated records and “what needs attention”.

Issues review queue

Quick views (mine, unassigned, high priority, came back, waiting 7+ days), filters, saved views, bulk review with preview and keyboard shortcuts.

Issue page

Why it matters, the recommended action, values at scan / now / expected, evidence versions, notes and history, and “re-check this rule now”.

Assessments and comparison

Run history with status, duration and policy fingerprint. Comparison refuses unlike-for-unlike runs, so improvements are never invented.

Coverage and readiness

Per business area: whether the source is ready, how many records it holds and which checks can run or are blocked.

Insights

Weekly flow of new, fixed and returning issues, where issues concentrate, recurring issues and how flagged records entered the ERP.

Duplicates

Candidate groups side by side with linked-document counts and an advisory “most used” record. It never merges anything.

Exceptions with four eyes

Accept a known observation for a limited time. One administrator requests, a different one approves, and it expires automatically.

Management report

Printable report per run: scope, sources read, a quality scorecard by dimension, results per check and exceptions.

Audit trail

Every event with time, user, before → after and reason: detected, assigned, exception approved, verified fixed, reopened.

Import check

Check a third-party CSV before importing it: missing names, bad e-mails, unknown countries and records that already exist in Dolibarr.

Alerts with a recipient firewall

Five alerts inside Dolibarr and/or by e-mail. They only ever reach active internal users and contain counts and links, never record values.

The 36 checks

What it checks

Fixed, documented and deterministic. Each check page lists its fields, priority, limits, why it matters and the recommended action.

Third parties · 12 checks

Blank name, e-mail syntax, missing customer/supplier code, country, address and payment terms; shared VAT, name + country and code candidates.

Products · 9 checks

Blank reference or label, shared reference and barcode, cost basis, missing unit, negative measurements, stray whitespace, stock that differs from warehouse totals.

Invoices and orders · 10 checks

Documents pointing to a missing third party or product, due date before invoice date, dates far in the future, net + tax ≠ gross, header ≠ lines.

Payments and stock · 3 checks

Payment allocations pointing to a missing invoice or exceeding the invoice total; stock rows pointing to a missing warehouse.

Bank accounts · 2 checks

IBAN checksum failures and one bank account shared by several third parties. Evidence is always masked.

Five quality dimensions

Every check belongs to one: completeness, validity, uniqueness, consistency or integrity, with a high, medium or low priority.

Every check ends in one honest state

No single “trust score”. Every percentage is shown with its counts: how many records were evaluated and how many were flagged.

Complete

The whole population was read. Only a complete run can close an issue.

Partial

A time, memory or row budget stopped the check. Never shown as clean, never extrapolated.

Not assessed

The source table or field is missing. Shown as such, not as a pass.

Screenshots

A tour of the module

Click any image to enlarge.

Data Trust & Governance: setup checklist, KPIs, trend and what needs attention
Dashboard — setup checklist, KPIs, trend and what needs attention
Data Trust & Governance: quick views, filters and the review queue
Issues — quick views, filters and the review queue
Data Trust & Governance: evidence at scan vs now, fix in Dolibarr, re-check
Issue page — evidence at scan vs now, fix in Dolibarr, re-check
Data Trust & Governance: candidate groups, never merged
Duplicates — candidate groups, never merged

Built for your team

Who it is for

Data quality is a team job. Each role gets a clear part of it.

Data steward

Works the issues queue: assigns, reviews, snoozes, dismisses with a reason, bulk-reviews and re-checks a rule after a fix.

Governance head / data owner

Reads the dashboard, scorecard, insights and management report, and approves exceptions.

Administrator

Chooses the policy and scan scope, schedules assessments, sets up alerts and retention, and manages rights.

Finance and accounting

Catches invoice totals that don't add up, broken payment allocations, invalid IBANs and shared bank accounts before they cost money.

Sales and purchasing

Keeps customer and supplier records complete and free of duplicates, and checks import files before they reach Dolibarr.

Auditors and management

Get evidence, owners, exceptions and a complete history instead of a one-off clean-up spreadsheet.

Trust & security

Safe by design

A governance tool has to be governed itself. These are the rules it follows.

Read-only on ERP data

It never edits, merges, blocks, deletes or repairs records. Its history lives in its own tables.

No invented numbers

No single trust score. Every percentage comes with its counts, partial runs are never shown as clean, and unlike runs are never compared.

Four-eyes exceptions

Exceptions are time-limited and need a second administrator to approve. Expired exceptions turn back into active issues.

Recipient firewall

Alerts only reach active internal users who can open Data Trust. Addresses stored on a customer, supplier, contact or member are refused. If in doubt, it doesn't send.

Nothing leaves your server

No AI, no telemetry, no outbound calls. Bank evidence is always masked, and e-mails carry counts and links only.

Full history

Every issue event is in the audit trail with who, when, before → after and why, exportable to CSV.

Quick start guide

From install to your first verified fix, without touching your data.

1

Download the module

Get the .zip file from DoliStore.

2

Install and enable

Go to Home → Setup → Modules/Applications, upload the zip and enable Data Trust & Governance. Its tables are created on enable.

3

Grant rights

View, Run, Review, Configure and Export. Give View deliberately, because issue evidence can contain names and e-mail addresses.

4

Choose a policy

Pick a business-type preset in Settings or on New assessment, and adjust any check.

5

Run your first assessment

Then open Coverage to see what each check could read and whether anything was partial or not assessed.

6

Work the issues

Assign owners, fix records in Dolibarr from the issue page, and use “Re-check this rule now” to confirm the fix.

7

Schedule and alert

Enable the daily assessment job and switch on the alerts you need, inside Dolibarr and/or by e-mail. Everything is off until you turn it on.

Frequently asked questions

Does it fix or merge my data automatically?

No. It never edits, merges, blocks or deletes ERP records. It shows the issue and the evidence, people fix the record in Dolibarr, and a later scan verifies the fix.

When is an issue considered fixed?

Only when a later complete assessment with the same policy shows the same record passing the same rule. Nobody can simply mark an issue as fixed.

Why is there no single trust score?

A single score hides what was actually checked. Every result is shown with its counts (records evaluated and flagged) and its coverage, so nothing is overstated.

Does it merge duplicate customers or products?

No. It shows candidate groups side by side with linked-document counts. A shared VAT number or reference is a prompt to look, not proof, because subsidiaries and variants can legitimately share one.

What if we accept a known issue?

Request a time-limited exception. A second administrator must approve it, and when it expires the issue becomes active again.

Can alerts reach customers or suppliers by mistake?

No. A recipient firewall allows only active internal users who can open Data Trust, and refuses any address that is also stored on a third party, contact or member.

Does it use AI or send data outside?

No. All 36 checks are deterministic. There is no AI, no telemetry and no outbound call unless you switch on an e-mail alert.

How large a database can it handle?

Scans run in batches within time and memory budgets. On synthetic tests, 200,000 third parties completed under a 128 MB PHP limit and 500,000 under 256 MB.

Which Dolibarr versions are supported?

Dolibarr 21, with PHP 7.4 or 8.2 and MySQL/MariaDB, on the current entity. MultiCompany sharing and PostgreSQL are not supported in this version.

Requirements

Runs on a standard Dolibarr 21 installation and reads the areas you have in use.

Dolibarr
21.x
PHP
7.4 – 8.2
Database
MySQL / MariaDB
Areas read
Third parties, products, invoices, orders, payments, stock, banks
Languages
EN · FR · DE · ES · IT
External services
None required

Data you can prove.
Fixes you can verify.

DoliStore listing and pricing will be published here at launch. Questions before then? Write to us.