---
description: "SAFe methodology core principles, round-table philosophy, search-first-reuse-always, evidence-based delivery. Apply to ALL work in this repository."
alwaysApply: true
---

# SAFe Core Principles

This repository uses **SAFe (Scaled Agile Framework) Agentic Workflow** with 11 specialized AI agent roles. You are part of a collaborative team where every contributor has equal voice.

## Round Table Philosophy

- Equal voice, mutual respect, shared responsibility
- Every agent and human contributor's input carries equal weight
- "Stop-the-line" authority: ANY contributor can halt work for architectural or security concerns
- Decisions are evidence-based, not authority-based

## Search First, Reuse Always

**MANDATORY before implementing anything:**

1. Search `patterns_library/` for existing code patterns
2. Search `specs/` for similar specifications
3. Search the codebase for similar implementations
4. Consult documentation (see CONTRIBUTING.md, docs/)
5. Only create new patterns if nothing exists (and that is BSA/Architect's job)

See `patterns_library/README.md` for the full pattern index (18+ patterns across 7 categories).

## SAFe Hierarchy

All work follows: **Epic > Feature > Story > Enabler**

1. BSA creates spec in `specs/{{TICKET_PREFIX}}-XXX-feature-spec.md`
2. System Architect validates the architectural approach
3. Implementation agents execute using discovered patterns
4. QAS validates against acceptance criteria
5. Evidence is attached to the project management ticket before POPM review

## Metacognitive Tags

Use these in specs to flag critical decisions:

- `#PATH_DECISION` -- Architectural path chosen (document alternatives)
- `#PLAN_UNCERTAINTY` -- Areas requiring validation
- `#EXPORT_CRITICAL` -- Security or compliance requirements

## Key Resources

- `CLAUDE.md` -- Full AI assistant context and development commands
- `AGENTS.md` -- All 11 agent roles, invocation patterns, exit states
- `CONTRIBUTING.md` -- Git workflow, commit standards, PR process
- `patterns_library/` -- Reusable code patterns (copy-paste ready)
- `docs/onboarding/` -- Setup guides and daily workflows
- `docs/guides/ROUND-TABLE-PHILOSOPHY.md` -- Collaboration principles

## Exit States

Every agent role has an explicit exit state defining its handoff point:

| Role             | Exit State                                |
|------------------|-------------------------------------------|
| BE-Developer     | "Ready for QAS"                           |
| FE-Developer     | "Ready for QAS"                           |
| Data-Engineer    | "Ready for QAS"                           |
| QAS              | "Approved for RTE"                        |
| RTE              | "Ready for HITL Review"                   |
| System Architect | "Stage 1 Approved - Ready for ARCHitect"  |
| HITL             | MERGED                                    |

## Quality Gates

| Gate            | Owner            | Blocking?                |
|-----------------|------------------|--------------------------|
| Stop-the-Line   | Implementer      | YES -- no AC = no work   |
| QAS Gate        | QAS              | YES -- no approval = stop|
| Stage 1 Review  | System Architect | YES -- pattern check     |
| Stage 2 Review  | ARCHitect-CLI    | YES -- architecture check|
| HITL Merge      | {{AUTHOR_NAME}}  | YES -- final authority   |

## Evidence-Based Delivery

- All work requires evidence attached to the project management ticket
- Evidence includes: command output, test results, validation screenshots
- No work is "done" without verifiable evidence