One local document pipeline

Editable Word handoff. Direct PDF. Evidence included.

Feed RusDox schema-backed YAML, JSON, TOML, or a designer-authored Word template. It returns a real DOCX, PDF bytes rendered from the same typed model, and an inspectable parity report—without Word, LibreOffice, LaTeX, or a cloud upload.

Best fit: recurring business documents that need both an editable Word handoff and a dependable PDF. Not a universal format converter or a full Word implementation.

Signed template registry

Start from a real Word design—with its receipts attached.

Browse curated invoices, proposals, reports, and compliance packs. Every entry includes its author, license, inputs, accessibility notes, output hashes, and inspectable DOCX/PDF parity evidence.

Featured verified template

Board and Compliance Report

A landscape board packet with repeated scorecard rows, conditional risk narrative, a visible footer, and verified parity evidence.

Browse verified templates

Install after verification

rusdox template add board-report
  • Manifest Ed25519 signature
  • Assets SHA-256 verified
  • Install Local and atomic
  • Core crate No template payload

Checksum-verified release

Install the binary. Leave user configuration untouched.

curl -fsSL https://raw.githubusercontent.com/OthmaneBlial/rusdox/2b3ca4eda4ab8389dc0e54198811bbaa3c368e44/scripts/install.sh | sh

Where the wedge matters

Keep editability, direct PDF output, and automated review in one system.

General converters, DOCX libraries, and PDF typesetters are excellent at different jobs. RusDox is deliberately narrower: dual-output business-document automation with proof that both artifacts contain the supported source semantics.

Structured specs

Developer-owned reports with a versioned schema.

Use readable YAML, JSON, or TOML for board packs, invoices, dashboards, and recurring exports. Validation points back to the source before rendering starts.

Word-native templates

Approved designs stay editable after automation.

Populate bounded placeholders, loops, and conditions in a designer-authored DOCX while preserving untouched package parts byte for byte.

Reviewable evidence

Make document parity visible in local work and CI.

Reopen DOCX, inspect PDF structure, compare semantic order, retain page snapshots, and add an optional visual baseline without sending content to a diff service.

Documentation and API

Learn the system, then automate it.

Read the authoring model, configuration, CLI surface, examples, and Rust API in one place.

Documentation workspace

Open the full docs and work from the real project material.

Browse getting started, YAML blocks, configuration, CLI usage, examples, Rust API guidance, and the community pages in one focused reading surface.

Go to docs

Key reading

Getting startedInstall the CLI and generate an editable DOCX, native PDF, and parity report.
Security reviewReview dated input limits, residual risks, and release supply-chain controls.

One model, two outputs

Author, render, then verify—without hiding the boundaries.

01

Choose structured data or a Word template

Specs suit developer-owned layouts. Word templates suit approved designs that must remain familiar and editable to non-developers.

blocks:
  - type: title
    text: Weekly Brief
  - type: body
    text: Pipeline grew 14% week over week.
02

Render both artifacts from one typed model

DOCX and PDF writers share validated content, metadata, page controls, and the documented layout subset. There is no hidden Office conversion step.

  • generated/*.docx
  • rendered/*.pdf
  • reports/*-parity.json
03

Verify the published contract

Green evidence covers the semantics RusDox promises. It never implies universal pixel identity across every Word or PDF viewer.

CLI DocumentSpec Studio Document

Examples

Real specs, real outputs, no mockups.

The cards below link directly to the YAML specs, DOCX files, and PDF previews.

Rendered preview of the RusDox Board Report
board packetmetricstables

Board Report

Leadership packet with narrative, metrics, and scorecard tables.

Rendered preview of the RusDox Executive Dashboard
dashboardKPIrisk

Executive Dashboard

KPI summary with narrative, delivery status, and risk tables.

Rendered preview of the RusDox Product Launch Brief
launchmilestonesreadiness

Product Launch Brief

Launch packet with milestones, checks, and customer-facing narrative.

Rendered preview of the RusDox Talent Profile
profileresumesignature

Talent Profile

Resume-style document combining narrative, tables, and a signature.

Rendered preview of the RusDox Invoice
invoicebillingtotals

Invoice

Compact billing layout with metadata, line items, and totals.

Rendered preview of the RusDox Meeting Notes
notesagendadecisions

Meeting Notes

Operational note with metadata, agenda bullets, and a decision log.

Contributor field guide

One fixture. Three acceptance checks. A contribution you can finish.

Ten maintained starter tasks remove the archaeology from a first pull request. Each issue points to a checked-in fixture, an isolated preparation command, and the evidence needed for review.

Start small

Choose a bounded, fixture-backed task.

Protocol, registry, documentation, and Windows integration tasks are ready to claim.

Open starter issues

Know the path

Trace authoring to both real outputs.

The architecture map shows where validation, OOXML, native PDF, parity, and adapters meet.

Read the architecture

Earn visible credit

Git-backed recognition, never invented names.

Meaningful merged work is credited in the contributor ledger, site, and relevant release notes.

Meet contributors

Reproducible performance evidence

Thirteen scenarios, explicit budgets, and raw host-separated results.

RusDox reproducible benchmark history across macOS arm64 and Linux x64
Linux x64 · 1,000-page dual output
1.40 s
macOS arm64 · 1,000-page dual output
1.57 s
Scenarios within reviewed budgets
13 / 13
Protocol date
2026-08-24

Each report records the CPU, OS, architecture, toolchain, input hash, output sizes, stage timings, and peak resident memory. Results from different machines are observations, not comparative scores.

Open performance budgets