Documentation · Export
PYD ENTERPRISES and Confluence
Publish a living specification space per legacy system: rule inventory, data dictionary, interface catalog and diagrams, regenerated on each approved scan.
Confluence
Documentation
- Documentation that is generated, not remembered
- Diagrams and formats
- Deployment notes
Documentation that is generated, not remembered
The usual discovery deliverable is a set of documents written by consultants from what they read and what retiring developers told them. It is out of date before the build phase starts. The Confluence export replaces that with pages PYD ENTERPRISES generates from the symbol graph and the approved rule set, so the documentation describes the code at a known commit or Endevor stage, and changes when the code does.
For each legacy system, PYD ENTERPRISES maintains a page tree in a space you nominate:
- System overview: inventory counts by language, entry points (CICS transactions, batch jobs, web endpoints), external interfaces and the scan reference.
- Rule inventory: one child page per DMN decision with its table, hit policy, inputs and outputs, and links from every row to source lines.
- Data dictionary: every record layout expanded from copybooks or DataContracts with offsets, lengths, PIC clauses, COMP-3 sizes and the JSON Schema type each maps to.
- Interface catalog: generated OpenAPI operations and AsyncAPI channels, with the legacy program or method each one came from.
Diagrams and formats
PYD ENTERPRISES writes pages in Confluence storage format. ER diagrams, sequence diagrams and dependency graphs are rendered to SVG attachments, with the Mermaid or PlantUML source attached beside them so your architects can edit the diagram outside Confluence. Tables are native Confluence tables, so readers can search and copy them; DMN XML and OpenAPI YAML are attached as files for anyone who wants the machine-readable form.
Regeneration only touches pages PYD ENTERPRISES owns. Each generated page carries a label and a short notice that it is generated, and PYD ENTERPRISES will not overwrite a page that a person has since edited by hand. Instead it publishes the new version as a child page and flags the conflict, so commentary your analysts add on a separate page or in comments is never lost.
Deployment notes
PYD ENTERPRISES publishes through Confluence's REST interface using a service account with add and edit permission on the target space. Confluence Data Center inside your network keeps traffic inside it. For Confluence Cloud, the PYD ENTERPRISES service needs an egress rule to your Atlassian site; decide with your security team whether source excerpts may appear on pages or whether pages should carry only line references. Air-gapped programs can export the same page tree as a static HTML bundle and import it through an approved route.
FAQ
Confluence questions.
Can we edit the generated pages?
You can, but PYD ENTERPRISES will then stop overwriting that page and publish its new version alongside it. Most teams keep generated pages read-only and add commentary on sibling pages or in comments.
Does Confluence render the Mermaid diagrams?
PYD ENTERPRISES attaches pre-rendered SVGs, so no diagram plugin is required. The Mermaid and PlantUML sources are attached beside them for editing.
How large does a space get for a big system?
Roughly one page per DMN decision plus one per major record layout and interface group. A system with several hundred decisions produces a few hundred pages, organized under a predictable tree.
RELATED INTEGRATIONS
See the Confluence flow on your own code.
Run PYD ENTERPRISES on 100,000 lines of COBOL or Java in your private VPC and review OpenAPI specs and DMN tables tomorrow.