How it works
From a set of figures to a finished report
How your data gets into bonoto
By hand
In the editor. A chapter is made of blocks: text, table, chart, image — and the data point that pulls a single figure out of the source.
From your spreadsheets
CSV lists and named Excel ranges. Files on SharePoint are read live and stay in sync. Several year columns become periods.
From the upstream system
Straight out of consolidation. bonoto does not replace your consolidation tool — it takes the figures it produces.
With your AI
Through the MCP server your agent writes straight into the source: create chapters, assign data points, carry text forward from last year’s wording and this year’s figures.
How a report is put together
Document
The report as a whole: title, languages, structure, state, output settings.
Section
A chapter with its own owner. A business unit maintains it, a state travels with it: draft → review → released.
Blocks
Text · table · chart · data point · Excel range · image. Tables and charts have editors of their own.
Glossary
Terms are explained once and linked throughout the report. Import from a list.
Collections
- Data Value
- ESRS Disclosure Requirement
- Glossary Term
- Report Document
- Report Section
- Glossar
- 1.3.1.1Basis for preparationGrundlage der Berichterstattung—29Published
- 1.3.1.1.1General basis for preparationAllgemeine Grundlage der AufstellungBP-130Published
- 1.3.1.2GovernanceUnternehmensführung—32Published
- 1.3.1.2.1The role of the administrative bodiesDie Rolle der VerwaltungsorganeGOV-133Published
- 1.3.1.3.2Interests and views of stakeholdersInteressen und Standpunkte der StakeholderSBM-240Changed
- 1.3.1.3.3Material impacts, risks and opportunitiesWesentliche Auswirkungen und RisikenSBM-341Changed
Blocks you drag into a section
- Text
- Data point
- Chart
- Table
- Image
- Excel range
Every section carries its number, both language versions, its mapping to the framework and its state. The blocks below are the same ones every output is built from.
Multilingual from the start
Every section, every block, every output in several languages. Today German and English; more are a matter of configuration.
The editor
Ready-made blocks.
Text, table, chart, image, data point and Excel range drag into a section. Every block knows what it is — which is why every output knows what to do with it.
Tables and charts have editors of their own. Anyone who can use Word needs no training.
- TEXT
- TABLE
- CHART
- DATA POINT1,842,9002026EUR k
The data point
One value, maintained once, identical everywhere in the report. What it means hangs on the value itself — not on its position in a table.
Everything is held as structured data, tables included. Whether you type them, paste them or link an Excel file: they become data, not formatting. Content and presentation are separate.
On top of that comes the mapping to the framework: external codes are mapped to internal data points in a maintainable table. As data, not as code — the catalog can be extended per reporting framework.
1,842,900
- Measure
- Revenue
- Period
- Fiscal year 2026
- Unit
- EUR thousand
- Entity
- Group total
- Origin
- Export 12 Feb 2027
- Framework
- mapped
Check it before anyone reads it
Language check
Terminology, spellings and style rules are checked against your editorial guide. Exceptions are set deliberately and stay visible.
Plausibility
Wrong units, thousand separators, factor-1000 slips and stale year references surface before they reach an output.
Closing
A closed report is frozen. Its outputs can be reproduced, even years later.
Why not a document
6 things Word and PDF cannot do
1 / 6
- 01
One content, many outputs
Word knows exactly one rendering: itself. In bonoto the content sits independent of any output. Web, print PDF, Word and XBRL come out side by side from the same state — not one after another by hand.
- 02
The figure knows its name
In Word, 1,842,900 is a piece of text. In bonoto it is a value with measure, year, unit, entity and origin. That is why it can go wherever it belongs — and changes everywhere at once.
- 03
No "final_final_v7"
Versions appear because everyone edits a copy. In bonoto there are no copies, only one state with history and release. Nobody has to merge comments and hope nothing was missed.
- 04
Structure instead of formatting
A heading 2 is a heading 2, not 18 point bold. Content and structure are cleanly separated in bonoto. How your report looks is set per output channel — once, not anew in every document.
- 05
Machine-readable from the start
Getting XBRL out of a PDF means reading a document backwards. In bonoto, XBRL and XML come from the same source as the print edition. Filing is not a project of its own, just another output channel.
- 06
The report knows its prior year
Comparatives are retyped today, and sooner or later someone mistypes. In bonoto the report knows which year it is and pulls the prior-year figures itself. What is right once stays right next cycle.
What comes out
Online report
As a public web version. Preview per section or as a whole.
Print PDF
A print view per language, with layout, fonts and page breaks.
Word
For copy editing and the agency: editable, cleaned up for tidy layout.
XML for InDesign
Including the mapping from markup to paragraph styles, and an import script.
XBRL
Machine-readable for registers, auditors and data recipients — structurally validated on every run.
XML for the commercial register
For filing with the register court — from the same state as every other output.
Single figures
As building blocks for your websites and presentations. Change the value and it changes there too.
Pick what you export: a section, a range or the whole report — not "all or nothing".
The report carries your identity, not ours.
MCP server
Agents are only as smart as the content they can read
The MCP server connects any agent straight to your report source in Strapi — structure, data points and the frameworks attached to them included.
The result is agents that do not just read your report, but understand it.
Agent requests access
Run access with
bonoto__read_chapter
- 1Yes, read only
- 2Yes, read and write
- 3No
The agent sees what the asking user is allowed to see. Every access is confirmed and logged.
The content platform where teams and AI build reports together
See the demo — 45 minutes, with your own figures. No slides.