feat(v2): agent consolidation — evaluator rename, tester agent, QA skill

.claude-plugin/marketplace.json

@@ -146,9 +146,9 @@
       ]
     },
     {
-      "name": "devflow-frontend-design",
-      "source": "./plugins/devflow-frontend-design",
-      "description": "Frontend design patterns - typography, color systems, spacing, motion, responsive design",
+      "name": "devflow-ui-design",
+      "source": "./plugins/devflow-ui-design",
+      "description": "UI design patterns - typography, color systems, spacing, motion, responsive design",
       "version": "1.8.3",
       "keywords": [
         "design",

.gitignore

@@ -20,7 +20,8 @@ plugins/*/agents/simplifier.md
 plugins/*/agents/coder.md
 plugins/*/agents/reviewer.md
 plugins/*/agents/resolver.md
-plugins/*/agents/shepherd.md
+plugins/*/agents/evaluator.md
+plugins/*/agents/tester.md
 plugins/*/agents/scrutinizer.md
 plugins/*/agents/validator.md
 npm-debug.log*

CLAUDE.md

@@ -28,7 +28,7 @@ Plugin marketplace with 17 plugins (8 core + 9 optional language/ecosystem), eac
 | `devflow-typescript` | TypeScript language patterns (optional) | No |
 | `devflow-react` | React framework patterns (optional) | No |
 | `devflow-accessibility` | Web accessibility patterns (optional) | No |
-| `devflow-frontend-design` | Frontend design patterns (optional) | No |
+| `devflow-ui-design` | UI design patterns (optional) | No |
 | `devflow-go` | Go language patterns (optional) | No |
 | `devflow-python` | Python language patterns (optional) | No |
 | `devflow-java` | Java language patterns (optional) | No |
@@ -50,8 +50,8 @@ Commands with Teams Variant ship as `{name}.md` (parallel subagents) and `{name}
 
 ```
 devflow/
-├── shared/skills/          # 37 skills (single source of truth)
-├── shared/agents/          # 10 shared agents (single source of truth)
+├── shared/skills/          # 38 skills (single source of truth)
+├── shared/agents/          # 11 shared agents (single source of truth)
 ├── plugins/devflow-*/      # 17 plugins (8 core + 9 optional language/ecosystem)
 ├── docs/reference/         # Detailed reference documentation
 ├── scripts/                # Helper scripts (statusline, docs-helpers)
@@ -130,20 +130,20 @@ Working memory files live in a dedicated `.memory/` directory:
 
 **Universal Skill Installation**: All skills from all plugins are always installed, regardless of plugin selection. Skills are tiny markdown files installed as `~/.claude/skills/devflow:{name}/` (namespaced to avoid collisions with other plugin ecosystems). Source directories in `shared/skills/` stay unprefixed — the `devflow:` prefix is applied at install-time only. Shadow overrides live at `~/.devflow/skills/{name}/` (unprefixed); when shadowed, the installer copies the user's version to the prefixed install target. Only commands and agents remain plugin-specific.
 
-**Model Strategy**: Explicit model assignments in agent frontmatter override the user's session model. Opus for analysis agents (reviewer, scrutinizer, shepherd), Sonnet for execution agents (coder, simplifier, resolver, skimmer), Haiku for I/O agents (git, synthesizer, validator).
+**Model Strategy**: Explicit model assignments in agent frontmatter override the user's session model. Opus for analysis agents (reviewer, scrutinizer, evaluator), Sonnet for execution agents (coder, simplifier, resolver, skimmer, tester), Haiku for I/O agents (git, synthesizer, validator).
 
 ## Agent & Command Roster
 
 **Orchestration commands** (spawn agents, never do agent work in main session):
 - `/specify` — Skimmer + Explore + Synthesizer + Plan + Synthesizer → GitHub issue
-- `/implement` — Git + Skimmer + Explore + Synthesizer + Plan + Synthesizer + Coder + Simplifier + Scrutinizer + Shepherd → PR
+- `/implement` — Git + Skimmer + Explore + Synthesizer + Plan + Synthesizer + Coder + Simplifier + Scrutinizer + Evaluator + Tester → PR
 - `/code-review` — 7-11 Reviewer agents + Git + Synthesizer
 - `/resolve` — N Resolver agents + Git
 - `/debug` — Agent Teams competing hypotheses
 - `/self-review` — Simplifier then Scrutinizer (sequential)
 - `/audit-claude` — CLAUDE.md audit (optional plugin)
 
-**Shared agents** (10): git, synthesizer, skimmer, simplifier, coder, reviewer, resolver, shepherd, scrutinizer, validator
+**Shared agents** (11): git, synthesizer, skimmer, simplifier, coder, reviewer, resolver, evaluator, tester, scrutinizer, validator
 
 **Plugin-specific agents** (1): claude-md-auditor
 

README.md

@@ -29,7 +29,8 @@ DevFlow: Ambient: IMPLEMENT/ORCHESTRATED
          → Validator: build ✓ typecheck ✓ lint ✓ tests ✓
          → Simplifier: cleaned up 3 files
          → Scrutinizer: 9-pillar quality check passed
-         → Shepherd: implementation matches request ✓
+         → Evaluator: implementation matches request ✓
+         → Tester: 5/5 QA scenarios passed ✓
 ```
 
 ```
@@ -47,7 +48,7 @@ DevFlow: Ambient: IMPLEMENT/ORCHESTRATED
 
 **18 parallel code reviewers.** Security, architecture, performance, complexity, consistency, regression, testing, and more. Each produces findings with severity, confidence scoring, and concrete fixes. Conditional reviewers activate when relevant (TypeScript for `.ts` files, database for schema changes). Every finding gets validated and resolved automatically.
 
-**34 skills grounded in expert material.** Every skill is backed by peer-reviewed papers, canonical books, and industry standards — security (OWASP, Shostack), architecture (Parnas, Evans, Fowler), performance (Brendan Gregg), testing (Beck, Meszaros), design (Wlaschin, Hickey). 200+ sources total.
+**38 skills grounded in expert material.** Every skill is backed by peer-reviewed papers, canonical books, and industry standards — security (OWASP, Shostack), architecture (Parnas, Evans, Fowler), performance (Brendan Gregg), testing (Beck, Meszaros), design (Wlaschin, Hickey). 200+ sources total.
 
 **Skill shadowing.** Override any built-in skill with your own version. Drop a file into `~/.devflow/skills/{name}/` and the installer uses yours instead of the default — same activation, your rules.
 
@@ -60,7 +61,7 @@ DevFlow: Ambient: IMPLEMENT/ORCHESTRATED
 ```
 devflow · feat/auth-middleware* · 3↑ · v1.8.3 +5 · 12 files · +234 -56
 Current Session ████░░░░ 42% · Session 5h ██░░░░░░ 18% · 7d █░░░░░░░ 8%
-Opus 4.6 [1m] · 23m · $1.24 · 2 CLAUDE.md · 4 MCPs · 8 hooks · 34 skills
+Opus 4.6 [1m] · 23m · $1.24 · 2 CLAUDE.md · 4 MCPs · 8 hooks · 38 skills
 ```
 
 **Security.** Deny lists block dangerous tool patterns out of the box — configurable during init.
@@ -88,7 +89,7 @@ See [docs/commands.md](docs/commands.md) for detailed usage.
 
 ## Language Support
 
-Optional plugins add language-specific patterns for TypeScript, React, Go, Python, Java, Rust, accessibility, and frontend design.
+Optional plugins add language-specific patterns for TypeScript, React, Go, Python, Java, Rust, accessibility, and UI design.
 
 ```bash
 npx devflow-kit init --plugin=typescript,react

docs/cli-reference.md

@@ -54,7 +54,7 @@ npx devflow-kit init --plugin=implement,code-review  # Install multiple
 | `devflow-typescript` | Language | TypeScript patterns |
 | `devflow-react` | Language | React patterns |
 | `devflow-accessibility` | Language | Web accessibility patterns |
-| `devflow-frontend-design` | Language | Frontend design patterns |
+| `devflow-ui-design` | Language | UI design patterns |
 | `devflow-go` | Language | Go patterns |
 | `devflow-python` | Language | Python patterns |
 | `devflow-java` | Language | Java patterns |

docs/commands.md

@@ -26,7 +26,8 @@ Executes a single task through the complete development lifecycle:
 4. **Implementation** — Write code on the feature branch
 5. **Validation** — Build, typecheck, lint, and test
 6. **Refinement** — Simplifier (code clarity) + Scrutinizer (9-pillar quality)
-7. **Alignment** — Shepherd verifies implementation matches the original request
+7. **Alignment** — Evaluator verifies implementation matches the original request
+8. **QA Testing** — Tester executes scenario-based acceptance tests
 
 Creates a PR when complete.
 

docs/reference/file-organization.md

@@ -9,13 +9,13 @@ devflow/
 ├── .claude-plugin/                   # Marketplace registry (repo root)
 │   └── marketplace.json
 ├── shared/
-│   ├── skills/                       # SINGLE SOURCE OF TRUTH (37 skills)
+│   ├── skills/                       # SINGLE SOURCE OF TRUTH (38 skills)
 │   │   ├── git/
 │   │   │   ├── SKILL.md
 │   │   │   └── references/
 │   │   ├── software-design/
 │   │   └── ...
-│   └── agents/                       # SINGLE SOURCE OF TRUTH (10 shared agents)
+│   └── agents/                       # SINGLE SOURCE OF TRUTH (11 shared agents)
 │       ├── git.md
 │       ├── synthesizer.md
 │       ├── coder.md
@@ -135,7 +135,7 @@ Skills and agents are **not duplicated** in git. Instead:
 
 ### Shared vs Plugin-Specific Agents
 
-- **Shared** (10): `git`, `synthesizer`, `skimmer`, `simplifier`, `coder`, `reviewer`, `resolver`, `shepherd`, `scrutinizer`, `validator`
+- **Shared** (11): `git`, `synthesizer`, `skimmer`, `simplifier`, `coder`, `reviewer`, `resolver`, `evaluator`, `tester`, `scrutinizer`, `validator`
 - **Plugin-specific** (1): `claude-md-auditor` — committed directly in its plugin
 
 ## Settings Override

docs/reference/skills-architecture.md

@@ -12,7 +12,7 @@ Shared patterns used by multiple agents.
 
 | Skill | Purpose | Used By |
 |-------|---------|---------|
-| `software-design` | Engineering patterns (Result types, DI, immutability, workaround labeling) | Coder, Scrutinizer, Resolver, Shepherd |
+| `software-design` | Engineering patterns (Result types, DI, immutability, workaround labeling) | Coder, Scrutinizer, Resolver, Evaluator |
 | `review-methodology` | 6-step review process, 3-category issue classification | Reviewer, Synthesizer |
 | `self-review` | 9-pillar self-review framework | Scrutinizer |
 | `docs-framework` | Documentation conventions (.docs/ structure, naming, templates) | Synthesizer |
@@ -21,6 +21,7 @@ Shared patterns used by multiple agents.
 | `agent-teams` | Agent Teams patterns for peer-to-peer collaboration, debate, consensus | /code-review, /implement, /debug |
 | `ambient-router` | Intent classification and proportional skill loading for ambient mode (unrestricted tools — orchestrator) | Ambient UserPromptSubmit hook |
 | `knowledge-persistence` | Record/load architectural decisions and pitfalls to `.memory/knowledge/` | /implement, /code-review, /resolve, /debug, /specify, /self-review |
+| `qa` | Scenario-based acceptance testing methodology, evidence collection | Tester |
 
 ### Tier 1b: Pattern Skills
 

plugins/devflow-ambient/.claude-plugin/plugin.json

@@ -20,7 +20,8 @@
     "validator",
     "simplifier",
     "scrutinizer",
-    "shepherd",
+    "evaluator",
+    "tester",
     "skimmer",
     "reviewer",
     "git",
@@ -48,6 +49,7 @@
     "documentation",
     "implementation-patterns",
     "knowledge-persistence",
+    "qa",
     "worktree-support"
   ]
 }

plugins/devflow-ambient/README.md

@@ -53,7 +53,7 @@ Skills are loaded via the Skill tool and work happens in the main session:
 
 | Intent | Pipeline |
 |--------|----------|
-| IMPLEMENT | Pre-flight → Coder → Validator → Simplifier → Scrutinizer → Shepherd |
+| IMPLEMENT | Pre-flight → Coder → Validator → Simplifier → Scrutinizer → Evaluator → Tester |
 | DEBUG | Hypotheses → parallel Explores → convergence → report → offer fix |
 | PLAN | Skimmer → Explores → Plan agent → gap validation |
 

plugins/devflow-implement/.claude-plugin/plugin.json

@@ -23,13 +23,15 @@
     "coder",
     "simplifier",
     "scrutinizer",
-    "shepherd",
+    "evaluator",
+    "tester",
     "validator"
   ],
   "skills": [
     "agent-teams",
     "implementation-patterns",
     "knowledge-persistence",
+    "qa",
     "self-review",
     "worktree-support"
   ]

plugins/devflow-implement/README.md

@@ -26,10 +26,11 @@ npx devflow-kit init --plugin=implement
 2. **Planning** - Plan agents design implementation approach
 3. **Implementation** - Coder agent implements on feature branch
 4. **Validation** - Validator runs build/test/lint checks
-5. **Self-Review** - Scrutinizer evaluates against 9-pillar framework
-6. **Alignment Check** - Shepherd validates against original request
-7. **Simplification** - Simplifier refines code clarity
-8. **PR Creation** - Git agent creates pull request
+5. **Simplification** - Simplifier refines code clarity
+6. **Self-Review** - Scrutinizer evaluates against 9-pillar framework
+7. **Alignment Check** - Evaluator validates against original request
+8. **QA Testing** - Tester executes scenario-based acceptance tests
+9. **PR Creation** - Git agent creates pull request
 
 ## Components
 
@@ -43,19 +44,17 @@ npx devflow-kit init --plugin=implement
 - `coder` - Autonomous implementation
 - `simplifier` - Code refinement
 - `scrutinizer` - Self-review (9-pillar framework)
-- `shepherd` - Alignment validation
+- `evaluator` - Alignment validation
+- `tester` - Scenario-based QA testing
 - `validator` - Build/test validation
 
-### Skills (9)
-- `software-design` - Result types, DI, immutability, workaround labeling
-- `git` - Git safety, atomic commits, PR descriptions
+### Skills (6)
+- `agent-teams` - Agent Teams orchestration patterns
 - `implementation-patterns` - CRUD, API, events
-- `testing` - Test quality, coverage
-- `boundary-validation` - Boundary validation
+- `knowledge-persistence` - Architectural decision recording
+- `qa` - Scenario-based acceptance testing
 - `self-review` - 9-pillar framework
-- `typescript` - TypeScript patterns
-- `react` - React patterns
-- `accessibility` - Keyboard, ARIA, focus management
+- `worktree-support` - Worktree-aware path resolution
 
 ## Output