sinapse-ai 1.24.0 → 1.25.0

package/.claude/agents/sinapse-trend-forecaster.md

@@ -0,0 +1,14 @@
+---
+name: trend-forecaster
+description: "Trend Forecasting Oracle"
+---
+
+# SINAPSE Claude Adapter: trend-forecaster
+
+Read and follow `squads/squad-research/agents/trend-forecaster.md` as the canonical source of truth.
+Adopt its persona, authority boundaries, activation protocol, declared dependencies,
+task inputs, outputs, gates and verification requirements.
+Work only within the authority declared by the canonical agent.
+Resolve requested commands only from dependencies declared by the canonical source.
+Use Claude Code native subagents or teams for delegation; never start a nested CLI.
+Follow project CLAUDE.md and the SINAPSE Constitution before acting.

package/.claude/agents/sinapse-ux-design-expert.md

@@ -0,0 +1,14 @@
+---
+name: ux-design-expert
+description: "Complete design workflow - user research, wireframes, design systems, token extraction, component building, and quality assurance"
+---
+
+# SINAPSE Claude Adapter: ux-design-expert
+
+Read and follow `.sinapse-ai/development/agents/ux-design-expert.md` as the canonical source of truth.
+Adopt its persona, authority boundaries, activation protocol, declared dependencies,
+task inputs, outputs, gates and verification requirements.
+Work only within the authority declared by the canonical agent.
+Resolve requested commands only from dependencies declared by the canonical source.
+Use Claude Code native subagents or teams for delegation; never start a nested CLI.
+Follow project CLAUDE.md and the SINAPSE Constitution before acting.

package/.claude/agents/sinapse-ux-designer.md

@@ -0,0 +1,14 @@
+---
+name: ux-designer
+description: "SINAPSE UX Design Expert autônomo. Frontend architecture, UI/UX design, wireframes, design system, accessibility, component design. 5 fases completas."
+---
+
+# SINAPSE Claude Adapter: ux-designer
+
+Read and follow `squads/squad-design/agents/ux-designer.md` as the canonical source of truth.
+Adopt its persona, authority boundaries, activation protocol, declared dependencies,
+task inputs, outputs, gates and verification requirements.
+Work only within the authority declared by the canonical agent.
+Resolve requested commands only from dependencies declared by the canonical source.
+Use Claude Code native subagents or teams for delegation; never start a nested CLI.
+Follow project CLAUDE.md and the SINAPSE Constitution before acting.

package/.claude/agents/sinapse-yvon-chouinard.md

@@ -0,0 +1,14 @@
+---
+name: yvon-chouinard
+description: "When you need to challenge growth-at-all-costs thinking. When deciding whether to sacrifice short-term profit for long-term integrity. When building a mission-driven company that puts purpose above shareholder returns. When evaluating..."
+---
+
+# SINAPSE Claude Adapter: yvon-chouinard
+
+Read and follow `squads/squad-council/agents/yvon-chouinard.md` as the canonical source of truth.
+Adopt its persona, authority boundaries, activation protocol, declared dependencies,
+task inputs, outputs, gates and verification requirements.
+Work only within the authority declared by the canonical agent.
+Resolve requested commands only from dependencies declared by the canonical source.
+Use Claude Code native subagents or teams for delegation; never start a nested CLI.
+Follow project CLAUDE.md and the SINAPSE Constitution before acting.

package/.claude/rules/software-engineering-applicability.md

@@ -0,0 +1,13 @@
+# Software Engineering Applicability
+
+Before selecting a project workflow, classify greenfield or brownfield,
+affected surfaces, data classes, deployment targets, reversibility and security,
+privacy and availability risk. Use the provider-neutral contract in
+`packages/engineering-applicability/index.js`; do not invent unknown signals.
+
+Treat QA as mandatory. Brownfield changes require characterization tests and an
+incremental change strategy. Frontend requires design-system grounding and
+desktop/mobile visual validation. Frontend/API requires application-security
+review. Database changes require migration, rollback and access-control review.
+Irreversible actions require explicit human approval. Full rationale:
+`docs/framework/software-engineering-applicability.md`.

package/.claude/skill-manifest.json

@@ -0,0 +1,41 @@
+{
+  "version": 1,
+  "skillIds": [
+    "sinapse",
+    "sinapse-agent",
+    "sinapse-analyst",
+    "sinapse-animations",
+    "sinapse-architect",
+    "sinapse-brand",
+    "sinapse-claude",
+    "sinapse-cloning",
+    "sinapse-commercial",
+    "sinapse-content",
+    "sinapse-copy",
+    "sinapse-council",
+    "sinapse-courses",
+    "sinapse-cyber",
+    "sinapse-data-engineer",
+    "sinapse-design",
+    "sinapse-dev",
+    "sinapse-devops",
+    "sinapse-finance",
+    "sinapse-growth",
+    "sinapse-loop",
+    "sinapse-orqx",
+    "sinapse-paidmedia",
+    "sinapse-pm",
+    "sinapse-po",
+    "sinapse-product",
+    "sinapse-qa",
+    "sinapse-research",
+    "sinapse-sm",
+    "sinapse-spec-driven",
+    "sinapse-squad-creator",
+    "sinapse-storytelling",
+    "sinapse-swarm",
+    "sinapse-ux-design-expert",
+    "snps",
+    "snps-orqx"
+  ]
+}

package/.claude/skills/api-review.md

@@ -0,0 +1,50 @@
+---
+name: api-review
+description: API and edge function review guidelines — activates when touching API routes or server functions
+paths: ["**/api/**", "**/routes/**", "supabase/functions/**", "pages/api/**", "app/api/**"]
+allowed-tools: [Read, Grep, Glob, Bash]
+user-invocable: false
+---
+
+# API & Edge Function Review
+
+When touching API routes or server functions, follow these guidelines:
+
+## Security (OWASP Top 10)
+- Validate ALL user input at the boundary (Zod preferred)
+- Never trust client-side data — re-validate on server
+- Sanitize output to prevent XSS
+- Use parameterized queries — never string concatenation for SQL
+- Implement rate limiting for public endpoints
+- Check authentication AND authorization on every protected route
+
+## Error Handling
+- Return consistent error shapes: `{ error: string, code: string, details?: unknown }`
+- Use appropriate HTTP status codes (400 client error, 500 server error)
+- Never expose internal error details to clients (stack traces, DB errors)
+- Log detailed errors server-side, return generic messages client-side
+
+## Request/Response
+- Validate request body with Zod schema
+- Type the response explicitly
+- Return early on validation failures (guard clauses)
+- Set appropriate CORS headers
+
+## Edge Functions (Supabase)
+- Import from `https://deno.land/` or `npm:` specifiers
+- Use `Deno.serve()` pattern
+- Always return `new Response()` — never throw unhandled
+- Set `Access-Control-Allow-Origin` in headers
+- Handle OPTIONS preflight requests
+
+## Performance
+- Implement caching where appropriate (Cache-Control headers)
+- Use database connection pooling
+- Paginate list endpoints (default limit: 20, max: 100)
+- Avoid N+1 queries — use JOINs or batch fetching
+
+## Documentation
+- Document endpoint in architecture docs BEFORE implementation
+- Include request/response examples
+- Document error codes and their meanings
+- Note rate limits and authentication requirements

package/.claude/skills/component-check.md

@@ -0,0 +1,47 @@
+---
+name: component-check
+description: Frontend component review guidelines — activates when touching React/JSX/TSX files
+paths: ["*.tsx", "*.jsx", "components/**", "pages/**", "app/**/*.tsx", "app/**/*.jsx"]
+allowed-tools: [Read, Grep, Glob]
+user-invocable: false
+---
+
+# Frontend Component Review
+
+When touching React components, follow these guidelines:
+
+## Component Structure
+- One component per file (exception: small internal sub-components)
+- Props interface defined and exported: `export interface FeatureProps { ... }`
+- Use functional components with hooks (never class components)
+- Explicit return type on component: `function Feature(props: FeatureProps): JSX.Element`
+
+## Naming
+- Components: PascalCase (`UserProfile`)
+- Files: kebab-case (`user-profile.tsx`)
+- Hooks: `use` prefix (`useUserProfile`)
+- Event handlers: `handle` prefix (`handleSubmit`)
+- Boolean props: `is`/`has`/`should` prefix (`isLoading`, `hasError`)
+
+## Imports
+- Always use absolute imports: `import { Button } from '@/components/ui/button'`
+- Never relative imports with `../../../`
+- Order: React → external → UI → utils → stores → features → CSS
+
+## Accessibility
+- Interactive elements must be keyboard accessible
+- Images need `alt` text
+- Form inputs need associated labels
+- Use semantic HTML (`nav`, `main`, `section`, `article`)
+
+## Performance
+- Memoize expensive computations: `useMemo`
+- Memoize callbacks passed to children: `useCallback`
+- Lazy load routes and heavy components: `React.lazy`
+- Avoid inline object/array literals in JSX props (causes re-renders)
+
+## State Management
+- Local state for UI-only state (`useState`)
+- Shared state via stores for cross-component state
+- Server state via data fetching hooks (SWR, React Query, or similar)
+- Never store derived state — compute it

package/.claude/skills/db-review.md

@@ -0,0 +1,41 @@
+---
+name: db-review
+description: Database schema review guidelines — activates automatically when touching schema files (Prisma, migrations, Supabase)
+paths: ["*.prisma", "migrations/**", "schema.*", "supabase/migrations/**", "supabase/functions/**"]
+allowed-tools: [Read, Grep, Glob, Bash]
+user-invocable: false
+---
+
+# Database Schema Review
+
+When touching database files, follow these guidelines from @data-engineer (Dara):
+
+## Schema Changes
+- Every schema change MUST have a corresponding migration
+- Never modify existing migrations — create new ones
+- Use descriptive migration names: `YYYYMMDDHHMMSS_description.sql`
+- Always include rollback logic (DOWN migration)
+
+## RLS Policies
+- Every new table MUST have Row Level Security enabled
+- Policies are per-table, not global — verify each table individually
+- Test policies with both authenticated and anonymous roles
+- Document policy intent in SQL comments
+
+## Query Optimization
+- Add indexes for columns used in WHERE, JOIN, ORDER BY
+- Use EXPLAIN ANALYZE before and after changes
+- Avoid N+1 queries — prefer JOINs or batch operations
+- Consider partial indexes for large tables with common filters
+
+## Naming Conventions
+- Tables: snake_case, plural (e.g., `user_profiles`)
+- Columns: snake_case (e.g., `created_at`)
+- Indexes: `idx_{table}_{column}` (e.g., `idx_users_email`)
+- Foreign keys: `fk_{table}_{ref_table}` (e.g., `fk_orders_users`)
+- Enums: snake_case, singular (e.g., `order_status`)
+
+## Before Committing
+- Run `supabase db diff` to verify migration matches schema
+- Check for breaking changes to existing RLS policies
+- Verify TypeScript types are regenerated if using typed client

package/.claude/skills/sinapse/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse
+description: Activate the SINAPSE supreme orchestrator in Claude Code.
+---
+
+# SINAPSE Claude Activation: sinapse-orqx
+
+Read `.sinapse-ai/development/agents/snps-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-sinapse-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-agent/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-agent
+description: Resolve and activate any canonical SINAPSE agent by ID in Claude Code.
+---
+
+# SINAPSE Parametric Agent Activator for Claude Code
+
+Use the agent ID supplied with this skill. Find the matching project subagent by its
+`name` frontmatter under `.claude/agents/sinapse-*.md`. Reject unknown IDs instead
+of guessing. Invoke that native subagent with `@<name>` when delegation is
+appropriate, and preserve its canonical authority and task contracts.

package/.claude/skills/sinapse-analyst/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-analyst
+description: Business Analyst (Scope). Use for market research, competitive analysis, user research, brainstorming session facilitation, structured ideation workshops, feasibility studies, i...
+---
+
+# SINAPSE Claude Activation: analyst
+
+Read `.sinapse-ai/development/agents/analyst.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-analyst` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-animations/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-animations
+description: Motion design, CSS animations, particulas, 3D, micro-interacoes
+---
+
+# SINAPSE Claude Activation: animations-orqx
+
+Read `squads/squad-animations/agents/animations-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-animations-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-architect/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-architect
+description: Architect (Stratum). Use for system architecture (fullstack, backend, frontend, infrastructure), technology stack selection (technical evaluation), API design (REST/GraphQL/tRPC...
+---
+
+# SINAPSE Claude Activation: architect
+
+Read `.sinapse-ai/development/agents/architect.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-architect` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-brand/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-brand
+description: Estrategia de marca, identidade visual, brandbook, auditoria de marca
+---
+
+# SINAPSE Claude Activation: brand-orqx
+
+Read `squads/squad-brand/agents/brand-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-brand-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-claude/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-claude
+description: Claude Code, MCP, hooks, skills, integracao avancada
+---
+
+# SINAPSE Claude Activation: claude-mastery-chief
+
+Read `squads/claude-code-mastery/agents/claude-mastery-chief.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-claude-mastery-chief` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-cloning/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-cloning
+description: Clonagem cognitiva, extracao de DNA, mind synthesis, knowledge bases
+---
+
+# SINAPSE Claude Activation: cloning-orqx
+
+Read `squads/squad-cloning/agents/cloning-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-cloning-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-commercial/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-commercial
+description: Vendas, funil, revenue ops, propostas comerciais, ofertas
+---
+
+# SINAPSE Claude Activation: commercial-orqx
+
+Read `squads/squad-commercial/agents/commercial-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-commercial-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-content/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-content
+description: Governanca editorial, calendario, adaptacao multi-plataforma, pipeline de conteudo
+---
+
+# SINAPSE Claude Activation: content-orqx
+
+Read `squads/squad-content/agents/content-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-content-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-copy/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-copy
+description: Copywriting persuasivo, headlines, landing pages, email sequences, conversao
+---
+
+# SINAPSE Claude Activation: copy-orqx
+
+Read `squads/squad-copy/agents/copy-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-copy-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-council/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-council
+description: Advisors estrategicos (Munger, Dalio, Thiel), mental models, decisoes
+---
+
+# SINAPSE Claude Activation: council-orqx
+
+Read `squads/squad-council/agents/council-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-council-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-courses/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-courses
+description: Curriculos, aulas, assessments, lancamento, producao educacional
+---
+
+# SINAPSE Claude Activation: courses-orqx
+
+Read `squads/squad-courses/agents/courses-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-courses-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-cyber/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-cyber
+description: Seguranca, threat intel, pentest, hardening, compliance
+---
+
+# SINAPSE Claude Activation: cyber-orqx
+
+Read `squads/squad-cybersecurity/agents/cyber-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-cyber-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-data-engineer/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-data-engineer
+description: Database Architect & Operations Engineer (Tensor). Use for database design, schema architecture, Supabase configuration, RLS policies, migrations, query optimization, data model...
+---
+
+# SINAPSE Claude Activation: data-engineer
+
+Read `.sinapse-ai/development/agents/data-engineer.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-data-engineer` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-design/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-design
+description: Design systems, componentes, tokens, acessibilidade, UX audit
+---
+
+# SINAPSE Claude Activation: design-orqx
+
+Read `squads/squad-design/agents/design-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-design-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-dev/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-dev
+description: Full Stack Developer (Pixel). Use for code implementation, debugging, refactoring, and development best practices
+---
+
+# SINAPSE Claude Activation: developer
+
+Read `.sinapse-ai/development/agents/developer.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-developer` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-devops/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-devops
+description: GitHub Repository Manager & DevOps Specialist (Pipeline). Use for repository operations, version management, CI/CD, quality gates, and GitHub push operations. ONLY agent authori...
+---
+
+# SINAPSE Claude Activation: devops
+
+Read `.sinapse-ai/development/agents/devops.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-devops` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-finance/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-finance
+description: Budget, pricing, rentabilidade, projecoes, analise financeira
+---
+
+# SINAPSE Claude Activation: finance-orqx
+
+Read `squads/squad-finance/agents/finance-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-finance-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-growth/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-growth
+description: Analytics, CRO, SEO, growth hacking, retencao, experimentacao
+---
+
+# SINAPSE Claude Activation: growth-orqx
+
+Read `squads/squad-growth/agents/growth-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-growth-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-loop/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-loop
+description: Run an explicitly requested bounded SINAPSE inspect-fix-verify loop with a completion criterion.
+---
+
+# SINAPSE Bounded Loop for Claude Code
+
+Activate only when the user explicitly requests this loop. Obtain a verifiable
+completion criterion, then repeat inspect -> delegate correction -> verify for at
+most three iterations. Stop on completion, block, user stop, or iteration three.
+Use Claude Code native subagents; never launch a nested Claude or Codex process.

package/.claude/skills/sinapse-orqx/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-orqx
+description: Route SINAPSE requests through the supreme ecosystem orchestrator and its specialist squads.
+---
+
+# SINAPSE Claude Activation: sinapse-orqx
+
+Read `.sinapse-ai/development/agents/snps-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-sinapse-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-paidmedia/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-paidmedia
+description: Meta Ads, Google Ads, campanhas, otimizacao, criativos, budget
+---
+
+# SINAPSE Claude Activation: paidmedia-orqx
+
+Read `squads/squad-paidmedia/agents/paidmedia-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-paidmedia-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-pm/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-pm
+description: Product Manager (Beacon). Use for PRD creation (greenfield and brownfield), epic creation and management, product strategy and vision, feature prioritization (MoSCoW, RICE), roa...
+---
+
+# SINAPSE Claude Activation: project-lead
+
+Read `.sinapse-ai/development/agents/project-lead.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-project-lead` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-po/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-po
+description: Product Owner (Axis). Use for backlog management, story refinement, acceptance criteria, sprint planning, and prioritization decisions
+---
+
+# SINAPSE Claude Activation: product-lead
+
+Read `.sinapse-ai/development/agents/product-lead.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-product-lead` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-product/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-product
+description: Product discovery, roadmap, MVP, analytics, estrategia de produto
+---
+
+# SINAPSE Claude Activation: product-orqx
+
+Read `squads/squad-product/agents/product-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-product-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-qa/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-qa
+description: Test Architect & Quality Advisor (Litmus). Use for comprehensive test architecture review, quality gate decisions, and code improvement. Provides thorough analysis including req...
+---
+
+# SINAPSE Claude Activation: quality-gate
+
+Read `.sinapse-ai/development/agents/quality-gate.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-quality-gate` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-research/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-research
+description: Market analysis, inteligencia competitiva, tendencias, pesquisa de audiencia
+---
+
+# SINAPSE Claude Activation: research-orqx
+
+Read `squads/squad-research/agents/research-orqx.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-research-orqx` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-sm/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-sm
+description: Scrum Master (Sync). Use for user story creation from PRD, story validation and completeness checking, acceptance criteria definition, story refinement, sprint planning, backlog...
+---
+
+# SINAPSE Claude Activation: sprint-lead
+
+Read `.sinapse-ai/development/agents/sprint-lead.md` as the canonical source of truth.
+Use the native Claude subagent `@agent-sprint-lead` for isolated delegation,
+or adopt the same persona in the current context when the workflow must remain here.
+Follow declared dependencies and authority boundaries. Never invoke a nested CLI.

package/.claude/skills/sinapse-spec-driven/SKILL.md

@@ -0,0 +1,11 @@
+---
+name: sinapse-spec-driven
+description: Prepare and coordinate SINAPSE PRD, epic, story, specification, and implementation-plan workflows in Codex.
+---
+
+# SINAPSE Spec-Driven Workflow for Claude Code
+
+Read `.sinapse-ai/development/workflows/spec-pipeline.yaml` and its referenced task
+contracts. Follow PRD -> Epic -> Story -> Validation -> Implementation, delegating
+each phase through Claude Code native subagents. A task contract controls inputs,
+outputs, gates and verification. Never implement before a validated Ready story.