SRC-RS: Sprint & Release Confidence Engine

SRC-RS is the simulation layer of ArchAI. It predicts the outcome of a sprint before the team begins working.

Simulation Pipeline

Step 1: Probability Aggregation

For every task in a WDP-TG plan, a completion probability is calculated: $P(task) = f(Effort, Risk, Dependencies, Experience)$ - High risk_level lowers probability. - Unresolved dependencies halt progress in the simulation until the parent task is "completed."

Step 2: Cumulative Simulation

The engine sequentially "runs" the sprint: - It respects team size and daily capacity. - It calculates an Epic Forecast based on the success rates of its child tasks. - It identifies Bottlenecks (tasks that block significant downstream work).

Step 3: Confidence Score

The final confidence_score (0-1) represents the mathematical probability that the entire release goals will be met.

Rationale & Recommendations

Every SRC-RS report includes Actionable Safety Recommendations linked to codebase evidence. - Example: "Task T001 has 40% probability due to missing tests in Module X. Add tests before sprint start."