How inthiiii/hyperreaility-sliit works
RE:L!NK is a proposed multi-agent decision-support application for planning telecom network restoration; the checked-in entry point is empty and the MVP is explicitly pending. The intended workflow loads incident and scenario data, coordinates specialist findings, builds and independently checks a constrained schedule, obtains human approval, then simulates approved actions and records outcomes. Material changes trigger renewed planning. The graph reflects documented plans, not verified runtime behavior; planned components without an implementing source file have null paths.
Main entry points: main.py Coordinator and state
Components 16
Evidence and investigation
- Scenario inputs
- Coordinator and state
project/main.py - Network Investigator
- Dependency analysis
Planning and review
- Field Resource Planner
- Service Priority Assessor
- Constrained scheduler
- Plan Reviewer
- Feasibility checker
Approval and simulation
- Operator clarification
- Human approval
- Action simulator
- Recorded outcomes
Evaluation and accountability
- State and trace storage
- Baseline evaluation
Other components
- Operations team
Connections 17
- Operations team → Scenario inputs: provides
README.md - Scenario inputs → Coordinator and state: loads and versions
README.md - Coordinator and state → Network Investigator: routes findings
README.md - Coordinator and state → Field Resource Planner: routes findings
README.md - Coordinator and state → Service Priority Assessor: routes findings
README.md - Coordinator and state → Plan Reviewer: routes candidate
README.md - Network Investigator → Dependency analysis: inspects dependencies
README.md - Coordinator and state → Constrained scheduler: builds schedule
README.md - Coordinator and state → Feasibility checker: checks plan
README.md - Plan Reviewer → Coordinator and state: requests revision
README.md - Coordinator and state → Operator clarification: escalates questions
README.md - Operator clarification → Coordinator and state: confirms updates
README.md - Operations team → Human approval: reviews plan
README.md - Human approval → Action simulator: authorizes actions
README.md - Action simulator → Recorded outcomes: records results
README.md - Coordinator and state → State and trace storage: versions state
README.md - Coordinator and state → Baseline evaluation: supports comparison
README.md