@polymorphism-tech/morph-spec 4.8.19 → 4.10.0

package/framework/rules/morph-workflow.md

@@ -1,86 +1,88 @@
-# MORPH-SPEC Workflow Rules
-
-> Always-active rules for all MORPH-SPEC managed projects.
-
----
-
-## Spec-First Mandate
-
-**NEVER skip to code without a specification.** Every feature must progress through phases:
-
-1. **Proposal** — User story + acceptance criteria (`0-proposal/proposal.md`)
-2. **Design** — Technical spec + C# contracts + decisions (`1-design/`)
-3. **Tasks** — Atomic task list (`3-tasks/tasks.md`)
-4. **Implement** — Code + recap (`4-implement/recap.md`)
-
-UI-heavy features also include **UI/UX** phase (`2-ui/`) between Design and Tasks.
-
----
-
-## Phase Commands
-
-| Command | Purpose |
-|---------|---------|
-| `/morph-proposal {feature}` | Full spec pipeline (phases 1–4, pauses for approval) |
-| `/morph-apply {feature}` | Implement feature (phase 5) |
-| `/morph-status` | Feature dashboard |
-
----
-
-## Output Paths
-
-All outputs go in `.morph/features/{feature}/`:
-
-| Phase | Path |
-|-------|------|
-| Proposal | `0-proposal/proposal.md` |
-| Spec | `1-design/spec.md` |
-| Contracts | `1-design/contracts.cs` |
-| Decisions | `1-design/decisions.md` |
-| UI Design | `2-ui/design-system.md`, `mockups.md`, `components.md`, `flows.md` |
-| Tasks | `3-tasks/tasks.md` |
-| Recap | `4-implement/recap.md` |
-
----
-
-## Checkpoints
-
-Run a checkpoint every 3 completed tasks:
-- Validate architecture compliance
-- Check package versions
-- Scan for security issues
-- Verify design system adherence (UI features)
-
----
-
-## Approval Gates
-
-**Design Gate:** Before moving from Design → Tasks, spec must be approved.
-**Implementation Gate:** Before starting implementation, task list must be approved.
-
-Check approval status: `morph-spec approval-status {feature}`
-
----
-
-## Protected Files
-
-**NEVER directly edit:**
-- `.morph/state.json` — Managed by CLI only
-- `.morph/framework/**` — Read-only framework content
-
----
-
-## Test File Policy
-
-When a test fails, always follow this order:
-
-1. **Analyze first** — determine if the IMPLEMENTATION is wrong or the TEST SPEC is wrong
-2. **Fix implementation first** — the test is the spec; trust it by default
-3. **Only modify a test file if the test expectation itself is incorrect** — wrong expected value, wrong behavior modeled
-4. **Before modifying any test file, explain WHY the test spec is wrong** — what the correct behavior is and why the test doesn't model it
-
-Do not modify test files to make a failing test pass when the implementation is the actual problem.
-
----
-
-*MORPH-SPEC by Polymorphism Tech*
+# MORPH-SPEC Workflow Rules
+
+> Always-active rules for all MORPH-SPEC managed projects.
+
+---
+
+## Spec-First Mandate
+
+**NEVER skip to code without a specification.** Every feature must progress through phases:
+
+1. **Proposal** — User story + acceptance criteria (`0-proposal/proposal.md`)
+2. **Design** — Technical spec + C# contracts + decisions (`1-design/`)
+3. **Plan** — Detailed implementation plan (`3-plan/plan.md`)
+4. **Tasks** — Atomic task list (`4-tasks/tasks.md`)
+5. **Implement** — Code + recap (`5-implement/recap.md`)
+
+UI-heavy features also include **UI/UX** phase (`2-ui/`) between Design and Tasks.
+
+---
+
+## Phase Commands
+
+| Command | Purpose |
+|---------|---------|
+| `/morph-proposal {feature}` | Full spec pipeline (phases 1–4, pauses for approval) |
+| `/morph-apply {feature}` | Implement feature (phase 5) |
+| `/morph-status` | Feature dashboard |
+
+---
+
+## Output Paths
+
+All outputs go in `.morph/features/{feature}/`:
+
+| Phase | Path |
+|-------|------|
+| Proposal | `0-proposal/proposal.md` |
+| Spec | `1-design/spec.md` |
+| Contracts | `1-design/contracts.cs` |
+| Decisions | `1-design/decisions.md` |
+| UI Design | `2-ui/design-system.md`, `mockups.md`, `components.md`, `flows.md` |
+| Plan | `3-plan/plan.md` |
+| Tasks | `4-tasks/tasks.md` |
+| Recap | `5-implement/recap.md` |
+
+---
+
+## Checkpoints
+
+Run a checkpoint every 3 completed tasks:
+- Validate architecture compliance
+- Check package versions
+- Scan for security issues
+- Verify design system adherence (UI features)
+
+---
+
+## Approval Gates
+
+**Design Gate:** Before moving from Design → Tasks, spec must be approved.
+**Implementation Gate:** Before starting implementation, task list must be approved.
+
+Check approval status: `morph-spec approval-status {feature}`
+
+---
+
+## Protected Files
+
+**NEVER directly edit:**
+- `.morph/state.json` — Managed by CLI only
+- `.morph/framework/**` — Read-only framework content
+
+---
+
+## Test File Policy
+
+When a test fails, always follow this order:
+
+1. **Analyze first** — determine if the IMPLEMENTATION is wrong or the TEST SPEC is wrong
+2. **Fix implementation first** — the test is the spec; trust it by default
+3. **Only modify a test file if the test expectation itself is incorrect** — wrong expected value, wrong behavior modeled
+4. **Before modifying any test file, explain WHY the test spec is wrong** — what the correct behavior is and why the test doesn't model it
+
+Do not modify test files to make a failing test pass when the implementation is the actual problem.
+
+---
+
+*MORPH-SPEC by Polymorphism Tech*

package/framework/skills/level-0-meta/mcp-registry.json

@@ -1,5 +1,5 @@
 {
-  "version": "2.0.0",
+  "version": "4.0.0",
   "description": "Centralized MCP registry with phase relevance and install configuration for morph-spec workflows",
   "mcps": {
     "context7": {
@@ -8,8 +8,16 @@
       "detection": "enabledPlugins includes context7 or mcpServers includes context7",
       "fallback": "WebSearch + WebFetch for library documentation",
       "usage": "Library documentation lookups, API reference, implementation patterns",
+      "healthCheck": {
+        "toolPattern": "context7",
+        "testCall": "Call resolve-library-id with libraryName 'react' and query 'react hooks'",
+        "testParams": { "libraryName": "react", "query": "react hooks" },
+        "successIndicator": "Returns a library ID (e.g., /facebook/react)",
+        "failureHint": "Context7 MCP not responding. Check that @upstash/context7-mcp is installed and the MCP server is running. Try restarting Claude Code."
+      },
       "install": {
         "autoInstall": true,
+        "transport": "stdio",
         "config": {
           "command": "cmd",
           "args": ["/c", "npx", "-y", "@upstash/context7-mcp@latest"]
@@ -27,8 +35,16 @@
       "detection": "mcpServers includes playwright",
       "fallback": "WebFetch for read-only page content, manual testing",
       "usage": "UI preview, smoke testing, page inspection, responsive testing, screenshot capture",
+      "healthCheck": {
+        "toolPattern": "playwright",
+        "testCall": "Call browser_navigate with url 'about:blank'",
+        "testParams": { "url": "about:blank" },
+        "successIndicator": "Page navigates successfully (no error thrown)",
+        "failureHint": "Playwright MCP not responding. Ensure @playwright/mcp is installed and browser binaries are available. Run 'npx playwright install' if needed."
+      },
       "install": {
         "autoInstall": true,
+        "transport": "stdio",
         "config": {
           "command": "cmd",
           "args": ["/c", "npx", "-y", "@playwright/mcp@latest"]
@@ -46,34 +62,25 @@
       "detection": "mcpServers includes supabase",
       "fallback": "Grep query files + Read type definitions for schema inference",
       "usage": "Schema analysis, migrations, RLS policies, data operations",
+      "healthCheck": {
+        "toolPattern": "supabase",
+        "testCall": "Call list_tables with no parameters",
+        "testParams": {},
+        "successIndicator": "Returns a list of tables (even if empty)",
+        "failureHint": "Supabase MCP not responding. If using remote mode, re-authenticate via /mcp in Claude Code. If using local mode, verify SUPABASE_URL and SUPABASE_SERVICE_ROLE_KEY in .claude/settings.local.json."
+      },
       "install": {
         "autoInstall": false,
+        "transport": "http",
         "config": {
-          "command": "cmd",
-          "args": ["/c", "npx", "-y", "@supabase/mcp-server-supabase"],
-          "env": {
-            "SUPABASE_URL": "",
-            "SUPABASE_SERVICE_ROLE_KEY": ""
-          }
+          "type": "http",
+          "url": "https://mcp.supabase.com/mcp"
         },
-        "credentials": [
-          {
-            "name": "Supabase Project URL",
-            "envVar": "SUPABASE_URL",
-            "secret": false,
-            "helpUrl": "https://supabase.com/dashboard/project/_/settings/api"
-          },
-          {
-            "name": "Service Role Key",
-            "envVar": "SUPABASE_SERVICE_ROLE_KEY",
-            "secret": true,
-            "helpUrl": "https://supabase.com/dashboard/project/_/settings/api"
-          }
-        ],
+        "credentials": [],
         "prerequisites": [],
         "stackRelevance": ["nextjs-supabase", "supabase", "nextjs"],
-        "warnings": ["NEVER connect to PRODUCTION databases. Use local/dev environment only."],
-        "setupGuide": null
+        "warnings": ["Uses OAuth — you will be prompted to authenticate via browser when first connecting.", "NEVER connect to PRODUCTION databases. Use local/dev environment only."],
+        "setupGuide": "Run: claude mcp add --transport http supabase https://mcp.supabase.com/mcp"
       }
     },
     "github": {
@@ -82,11 +89,19 @@
       "detection": "mcpServers includes github",
       "fallback": "Bash `gh` CLI commands",
       "usage": "Repo metadata, code search, issue creation, PR management",
+      "healthCheck": {
+        "toolPattern": "github",
+        "testCall": "Call any read-only GitHub tool (e.g., search_repositories or get_repo) to verify connectivity",
+        "testParams": {},
+        "successIndicator": "Returns repository data or search results without auth errors",
+        "failureHint": "GitHub MCP not responding. If using Docker, ensure Docker Desktop is running. Verify GITHUB_PERSONAL_ACCESS_TOKEN env var is set. Ensure the token has not expired and has the required scopes (repo, read:org)."
+      },
       "install": {
         "autoInstall": false,
+        "transport": "stdio",
         "config": {
-          "command": "cmd",
-          "args": ["/c", "npx", "-y", "@modelcontextprotocol/server-github"],
+          "command": "docker",
+          "args": ["run", "-i", "--rm", "-e", "GITHUB_PERSONAL_ACCESS_TOKEN", "ghcr.io/github/github-mcp-server"],
           "env": {
             "GITHUB_PERSONAL_ACCESS_TOKEN": ""
           }
@@ -99,39 +114,43 @@
             "helpUrl": "https://github.com/settings/tokens"
           }
         ],
-        "prerequisites": [],
+        "prerequisites": [
+          {
+            "name": "Docker",
+            "checkCommand": "docker --version",
+            "installUrl": "https://docs.docker.com/get-docker/"
+          }
+        ],
         "stackRelevance": ["*"],
         "warnings": [],
-        "setupGuide": null
+        "setupGuide": "Official GitHub MCP Server (github/github-mcp-server). Requires Docker."
       }
     },
-    "figma": {
-      "phases": ["uiux"],
-      "tools": ["get_file"],
-      "detection": "mcpServers includes figma",
-      "fallback": "Read CSS/SCSS files for existing design tokens",
-      "usage": "Design token extraction, component specs, color/typography from Figma files",
+    "vercel": {
+      "phases": ["implement"],
+      "tools": ["search_docs", "list_projects", "get_deployments", "get_deployment_logs", "manage_env_vars"],
+      "detection": "mcpServers includes vercel",
+      "fallback": "Bash `vercel` CLI commands",
+      "usage": "Deployment monitoring, project management, environment variables, log analysis",
+      "healthCheck": {
+        "toolPattern": "vercel",
+        "testCall": "Call search_docs with a query to verify connectivity",
+        "testParams": { "query": "deployments" },
+        "successIndicator": "Returns documentation search results without auth errors",
+        "failureHint": "Vercel MCP not responding. Re-authenticate via /mcp in Claude Code. Ensure you have authorized the Vercel OAuth connection."
+      },
       "install": {
         "autoInstall": false,
+        "transport": "http",
         "config": {
-          "command": "cmd",
-          "args": ["/c", "npx", "-y", "@anthropic/mcp-server-figma"],
-          "env": {
-            "FIGMA_ACCESS_TOKEN": ""
-          }
+          "type": "http",
+          "url": "https://mcp.vercel.com"
         },
-        "credentials": [
-          {
-            "name": "Figma Access Token",
-            "envVar": "FIGMA_ACCESS_TOKEN",
-            "secret": true,
-            "helpUrl": "https://www.figma.com/developers/api#access-tokens"
-          }
-        ],
+        "credentials": [],
         "prerequisites": [],
-        "stackRelevance": ["blazor", "nextjs", "react", "vue"],
-        "warnings": [],
-        "setupGuide": null
+        "stackRelevance": ["nextjs", "nextjs-supabase", "vercel"],
+        "warnings": ["Uses OAuth — you will be prompted to authenticate via browser when first connecting."],
+        "setupGuide": "Run: claude mcp add --transport http vercel https://mcp.vercel.com"
       }
     },
     "docker": {
@@ -140,8 +159,16 @@
       "detection": "mcpServers includes docker",
       "fallback": "Bash `docker` CLI commands",
       "usage": "Container management, build and run containers",
+      "healthCheck": {
+        "toolPattern": "docker",
+        "testCall": "Call any read-only Docker tool to verify Docker daemon connectivity",
+        "testParams": {},
+        "successIndicator": "Tool responds without connection errors (Docker daemon is reachable)",
+        "failureHint": "Docker MCP not responding. Ensure Docker Desktop is running and the Docker daemon is accessible. Run 'docker info' to verify."
+      },
       "install": {
         "autoInstall": false,
+        "transport": "stdio",
         "config": {
           "command": "cmd",
           "args": ["/c", "npx", "-y", "@modelcontextprotocol/server-docker"]
@@ -165,8 +192,16 @@
       "detection": "mcpServers includes azure",
       "fallback": "Bash `az` CLI commands",
       "usage": "Cloud resource management, deployment",
+      "healthCheck": {
+        "toolPattern": "azure",
+        "testCall": "Call any read-only Azure tool to verify credentials and connectivity",
+        "testParams": {},
+        "successIndicator": "Tool responds without authentication or subscription errors",
+        "failureHint": "Azure MCP not responding. Verify AZURE_SUBSCRIPTION_ID in .claude/settings.local.json and ensure 'az login' has been run. Check that Azure CLI is installed with 'az --version'."
+      },
       "install": {
         "autoInstall": false,
+        "transport": "stdio",
         "config": {
           "command": "cmd",
           "args": ["/c", "npx", "-y", "@azure/mcp-server"],
@@ -198,10 +233,10 @@
   "phaseMatrix": {
     "proposal": ["context7"],
     "setup": ["github"],
-    "uiux": ["playwright", "figma", "context7"],
+    "uiux": ["playwright", "context7"],
     "design": ["supabase", "context7", "github"],
     "clarify": ["context7", "github"],
     "tasks": ["context7", "github"],
-    "implement": ["supabase", "context7", "playwright", "github", "docker", "azure"]
+    "implement": ["supabase", "context7", "playwright", "github", "vercel", "docker", "azure"]
   }
 }

package/framework/skills/level-0-meta/{brainstorming → morph-brainstorming}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: brainstorming
+name: morph:brainstorming
 description: Morph-spec-aware brainstorming that loads project context, explores multiple design approaches, asks clarifying questions, and produces proposal.md or decisions.md. Use before designing a feature, when facing architectural decisions with multiple valid approaches, or when a feature needs requirements exploration before committing to a direction.
 user-invocable: true
 argument-hint: "[feature-name or topic]"
@@ -53,26 +53,23 @@ Read: .morph/features/{feature}/0-proposal/proposal.md
 
 ### Step 2: Ask Clarifying Questions
 
-Generate **3-5 focused questions** for the user:
+Use `AskUserQuestion` to collect responses — **NEVER list questions as plain text**.
 
-```markdown
-### Q1: Scope
-What is the minimum viable version of this feature?
-
-### Q2: Constraints
-Are there performance/budget/timeline constraints?
+- Máximo 4 perguntas por chamada. Se precisar de 5, faça 2 chamadas sequenciais (Q1-Q4 → Q5).
+- Para cada pergunta, defina 2-4 opções representativas (o usuário pode usar "Other" para resposta livre).
+- **Aguarde o retorno do tool antes de prosseguir para o Step 3.**
 
-### Q3: Integration
-Does this need to integrate with existing features?
+Perguntas-padrão a adaptar para o contexto da feature:
 
-### Q4: Users
-Who are the primary users of this feature?
-
-### Q5: Success Criteria
-How will we know this feature is working correctly?
-```
+| Header (≤12 chars) | Pergunta | Opções sugeridas |
+|--------------------|----------|------------------|
+| `Escopo` | Qual é a versão mínima viável desta feature? | MVP básico / Versão completa / Em aberto |
+| `Constraints` | Há restrições de performance, budget ou prazo? | Sem restrições / Performance crítica / Budget limitado |
+| `Integração` | Precisa integrar com features ou sistemas existentes? | Sim, integra com X / Feature isolada / A definir |
+| `Usuários` | Quem são os usuários primários? | End-users / Admins / Sistema interno |
+| `Sucesso` | Como saberemos que a feature funciona corretamente? | Métricas definidas / Testes de aceitação / Feedback do usuário |
 
-**Wait for user responses before proceeding.**
+> Adapte as opções ao contexto real da feature — as acima são ponto de partida, não template fixo.
 
 ### Step 3: Generate 2-3 Approaches
 

package/framework/skills/level-0-meta/morph-checklist/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: morph-checklist
+name: morph:checklist
 description: Pre-deploy, security, SEO, performance, accessibility, and LGPD compliance checklists for MORPH-SPEC projects. Use before deploying to production, during security audits, when optimizing for SEO or performance, or when validating Brazilian LGPD data protection compliance.
 user-invocable: true
 allowed-tools: Read, Write, Edit, Bash, Glob, Grep
@@ -7,7 +7,7 @@ allowed-tools: Read, Write, Edit, Bash, Glob, Grep
 
 # MORPH Checklists
 
-Types: `deploy`, `security`, `seo`, `performance`, `accessibility`, `legal-brazil`, `simulation` (ver skill: `simulation-checklist`)
+Types: `deploy`, `security`, `seo`, `performance`, `accessibility`, `legal-brazil`, `simulation` (ver skill: `morph:simulation-checklist`)
 
 ---
 

package/framework/skills/level-0-meta/{code-review → morph-code-review}/SKILL.md

@@ -1,5 +1,5 @@
 ---
-name: code-review
+name: morph:code-review
 description: .NET/C# code review checklist covering naming conventions, architecture layer integrity, async patterns, logging, error handling, DI lifetimes, and DTO contracts. Use after implementing .NET code, before creating PRs, or when reviewing C# code for compliance with MORPH-SPEC standards.
 user-invocable: true
 allowed-tools: Read, Write, Edit, Bash, Glob, Grep
@@ -26,7 +26,7 @@ git diff --name-only main...HEAD -- "*.cs" | wc -l
 **Pular revisão se:**
 - 0 arquivos `.cs` alterados → sem código backend para revisar
 - Apenas mudanças em `*.json`, `*.csproj`, `*.md`, arquivos de migration → escopo de infra
-- Já existe `.morph/features/$ARGUMENTS/4-implement/code-review.md` criado hoje → já revisado
+- Já existe `.morph/features/$ARGUMENTS/5-implement/code-review.md` criado hoje → já revisado
 
 Se skip: output `"Skipping code-review: [motivo]"` e parar.