References

This section collects the reference material of the documentation: reference documents and requirements material, API and MCP references, and the glossary. The system diagrams are collected in the Diagram Gallery.

Reference Documents and Artifacts

• This documentation — source repository: https://github.com/eic/epic-wfms-docs
• ePIC Distributed Workflow Management System requirements: https://www.overleaf.com/project/67bdf89a3d44a138da503dea
• The ePIC Streaming Computing Model: https://zenodo.org/records/14675920
• ePIC Production System: Status and Plans (ePIC S&C meeting talk, June 24 2026 — the definitive current (June 2026) description of the production system, with interface views and near-term plans): https://docs.google.com/presentation/d/1Mhc5Isfq0dKOqYbf7fdFQvu7tM86gsKLYx542aYVL0U/edit
• Production bot demo video (June 2026): https://drive.google.com/file/d/1VXIJJHfCaqE2iK_5QawJNamciGhFVQEI/view
• PanDA documentation: https://panda-wms.readthedocs.io/
• PanDA paper: https://link.springer.com/article/10.1007/s41781-024-00114-3
• Rucio paper: https://link.springer.com/article/10.1007/s41781-019-0026-3
• iDDS paper: https://link.springer.com/article/10.1140/epjc/s10052-025-15275-7
• EIC/ePIC Software Statement of Principles: https://eic.github.io/activities/principles.html
• Streaming workflow testbed architecture: https://github.com/BNLNPPS/swf-testbed/blob/main/docs/architecture.md
• Testbed progress notes (working document — meeting notes, progress highlights, technical detail): https://docs.google.com/document/d/1PUoo-W6dCeOKsD4VubYTgSxBHBUb4D5dYfVy1oLYh7E/edit
• epicprod implementation repository: https://github.com/BNLNPPS/swf-monitor
• epicprod production domain repository (production applications and documentation consolidate here; its docs/ARCHITECTURE_MAP.md records what lives where): https://github.com/BNLNPPS/swf-epicprod
• PCS — Physics Configuration System: https://github.com/BNLNPPS/swf-epicprod/blob/main/docs/PCS.md
• epicprod task catalog design: https://github.com/BNLNPPS/swf-epicprod/blob/main/docs/EPICPROD_TASK_CATALOG.md
• Production operations runbook: https://github.com/BNLNPPS/swf-epicprod/blob/main/docs/EPICPROD_OPS.md
• Production operations agent design: https://github.com/BNLNPPS/swf-epicprod/blob/main/docs/EPICPROD_OPS_AGENT.md
• Validation integration plan: https://github.com/BNLNPPS/swf-epicprod/blob/main/docs/EPICPROD_VALIDATION.md
• Production request questionnaire design: https://github.com/BNLNPPS/swf-epicprod/blob/main/docs/EPICPROD_QUESTIONNAIRE.md
• argus-ai assessment application design: https://github.com/BNLNPPS/corun-ai/blob/master/docs/argus-ai.md

API and MCP References

The programmatic interfaces are documented with the implementation:

• swf-monitor REST API reference: https://github.com/BNLNPPS/swf-monitor/blob/main/docs/API_REFERENCE.md
• swf-monitor MCP server overview: https://github.com/BNLNPPS/swf-monitor/blob/main/docs/MCP.md
• MCP tool catalog: https://github.com/BNLNPPS/swf-monitor/blob/main/docs/MCP_TOOL_REFERENCE.md
• MCP client setup: https://github.com/BNLNPPS/swf-monitor/blob/main/docs/MCP_CLIENTS.md
• PanDA client and API documentation: https://panda-wms.readthedocs.io/en/latest/client/client.html

Glossary

• DDM: Distributed data management
• Echelon: The ePIC computing model's resource hierarchy: E0, E1, E2, and E3
• EVGEN: Event-generation data, the generator outputs consumed as production inputs
• Harvester: PanDA service that manages worker submission and execution at computing sites
• iDDS: intelligent Data Delivery Service, the workflow orchestration component of the PanDA ecosystem
• JEDI: The PanDA task definition and execution layer
• MCP: Model Context Protocol, the structured tool interface serving AI clients
• PanDA: Production and Distributed Analysis system used for distributed workload management
• PCS: Physics Configuration System, the epicprod subsystem for production configuration and campaign records
• RECO: Reconstruction output data products
• Rucio: Distributed data management system used for dataset, replica, and data-placement management
• SSE: Server-Sent Events, the push channel delivering asynchronous notifications to browser pages
• STF: Super time frame, an aggregate unit of streaming detector data used in post-DAQ processing
• TF: Time frame, a time-slice unit of streaming detector data
• WFMS: Workflow Management System