What depends on what, and can we go back: the toolkit dependency tree of every named application snapshot (toolkit, toolkit snapshot, system flag) as a table and an SVG tree - read from the product database because /ops exposes no dependencies - and a rollback advisor per container: every version with its running instances (/ops count), its open tasks (search), the active / default / tip flags and an advice line (drain or migrate the current version, re-activate an earlier idle version, archive an idle one).
Target: IBM BAW 20.0.0.1 and later (traditional) and CP4BA; needs the Operations REST API (/ops) and, for the database-backed tabs, the engine datasource. Built from out-of-the-box building blocks only (System Data, UI Toolkit, one client-side human service,
server-side service flows calling the product REST API as a technical user; where no API exists, a read-only SELECT on the product
database through the engine's own datasource) - no third-party toolkit, no custom Java.
- Import
packages/Snapshot-Lifecycle-1.0.twx(Process Center / Workflow Center console: Import Process App; CP4BA: Business Automation Studio > Import). - Set the environment variables of the snapshot (below): the server base URL as seen from the server itself and the technical user the flows call the REST API with. The package ships no password.
- Expose the dashboard to a team (the package's default team holds
celladmin; change it) and open it from Process Portal or the playback URL of the human service.
| Tab | Content |
|---|---|
| Dependencies | application filter, all / active snapshots, Dependencies, Export CSV; SVG tree; table |
| Rollback advisor | container acronym, Advise; version table with advice |
serverBaseURL, restAuthUser, restAuthPassword, restTrustAllCertificates, appTitle, opsLoginPath, opsBasePath, dbJndiName.
- docs/DESIGN.md - what the app does, every service flow with its REST calls and SQL, business objects, build and test procedure, pitfalls met on the way.
- docs/RECIPES-AND-PITFALLS.md - the recipe collection of all the kit apps: shared helpers, verified search / team / instance / operations / product database facts, build and deployment pitfalls.
The application is generated by a small Python program on the twxkit generator (standard library only, no base export). The generator
and the kit are served to AI agents and developers by the BAW Knowledge MCP server (https://bawmcp.dosvak.com/mcp, see
dosvak/baw-mcp): get_package('snapshot lifecycle (snaplc)') returns this package, get_tool('build_snaplc.py')
the generator, get_topic('howto-kit-utility-app-recipes') the recipes.
Apache License 2.0 - see LICENSE and NOTICE. IBM, IBM Business Automation Workflow, IBM Business Process Manager and IBM Cloud Pak are trademarks of International Business Machines Corporation; this project is not affiliated with or supported by IBM.