Add quality gates, role library, and templates for team lifecycle v3

- Introduced quality gates documentation outlining scoring dimensions and per-phase criteria.
- Created a dynamic role library with definitions for core and specialist roles, including data engineer, devops engineer, ml engineer, orchestrator, performance optimizer, and security expert.
- Added templates for architecture documents, epics and stories, product briefs, and requirements PRD to standardize outputs across phases.

.claude/skills/team-lifecycle-v3/specs/artifact-contract-spec.md

@@ -0,0 +1,101 @@
+# Artifact Contract Specification
+
+## Overview
+
+Every worker role must generate an artifact manifest in Phase 4 to enable quality gating and automatic artifact discovery.
+
+## Manifest Schema
+
+**Location**: `<session>/artifacts/<task-id>/artifact-manifest.json`
+
+```json
+{
+  "artifact_id": "uuid-...",
+  "creator_role": "role-name",
+  "artifact_type": "spec | architecture | plan | code | test | review",
+  "version": "1.0.0",
+  "path": "./artifacts/output.md",
+  "dependencies": ["upstream-artifact-id"],
+  "validation_status": "pending | passed | failed",
+  "validation_summary": "Details...",
+  "metadata": {
+    "complexity": "low | medium | high",
+    "priority": "P0 | P1 | P2"
+  }
+}
+```
+
+## Field Definitions
+
+| Field | Type | Description |
+|-------|------|-------------|
+| artifact_id | string | UUID for this artifact |
+| creator_role | string | Role that created this artifact |
+| artifact_type | enum | Type of artifact (spec/architecture/plan/code/test/review) |
+| version | string | Semantic version |
+| path | string | Relative path to artifact file |
+| dependencies | array | List of upstream artifact IDs this depends on |
+| validation_status | enum | pending/passed/failed |
+| validation_summary | string | Human-readable validation result |
+| metadata | object | Additional context (complexity, priority, etc.) |
+
+## Validation Status
+
+- **pending**: Artifact created, validation not yet run
+- **passed**: All quality checks passed, ready for downstream consumption
+- **failed**: Quality issues detected, blocks downstream spawn
+
+## Coordinator Integration
+
+The coordinator checks manifests in `handleCallback`:
+
+1. Read `<session>/artifacts/<task-id>/artifact-manifest.json`
+2. If `validation_status == "passed"`:
+   - Register to artifact_registry
+   - Spawn next dependent task
+3. If `validation_status == "failed"`:
+   - Create fix task
+   - Notify user
+   - Block downstream tasks
+
+## Context Artifacts Discovery
+
+Workers read `context-artifacts.json` to discover upstream artifacts:
+
+```json
+{
+  "artifacts": [
+    {
+      "artifact_id": "uuid-...",
+      "artifact_type": "spec",
+      "path": "./spec/discovery-context.json",
+      "creator_role": "analyst"
+    }
+  ]
+}
+```
+
+## Auto-Discovery (v5)
+
+In v5, workers declare `input_artifact_types: []` in frontmatter. The coordinator automatically discovers and provides matching artifacts without manual path configuration.
+
+## Quality Gates
+
+Validation checks vary by artifact type:
+
+| Artifact Type | Required Checks |
+|---------------|----------------|
+| spec | Completeness, schema compliance |
+| architecture | ADR presence, component diagram |
+| plan | Task count (2-7), dependency graph validity |
+| code | Syntax check, test coverage |
+| test | Test count > 0, all passing |
+| review | Issue count, severity distribution |
+
+## Example Workflow
+
+1. **Analyst** creates spec → manifest with `validation_status: "passed"`
+2. **Coordinator** reads manifest → registers to artifact_registry
+3. **Writer** spawned → reads `context-artifacts.json` → finds analyst's spec
+4. **Writer** creates document → manifest with `validation_status: "passed"`
+5. **Coordinator** spawns next role based on dependency graph