@kal-elsam/kairo-runtime 0.1.0

package/repo-template/.cursor/commands/sdd-intake.md

@@ -0,0 +1,18 @@
+# /sdd-intake
+
+Analyze requirement, constraints, assumptions and missing context. Output a concise intake brief.
+
+## Source
+
+- `AGENTS.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/harness.md`
+
+## Output
+
+```txt
+Result:
+Files:
+Risks:
+Next action:
+```

package/repo-template/.cursor/commands/sdd-plan.md

@@ -0,0 +1,18 @@
+# /sdd-plan
+
+Create technical plan from approved spec.
+
+## Source
+
+- `AGENTS.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/harness.md`
+
+## Output
+
+```txt
+Result:
+Files:
+Risks:
+Next action:
+```

package/repo-template/.cursor/commands/sdd-spec.md

@@ -0,0 +1,18 @@
+# /sdd-spec
+
+Create or update the spec in docs/specs based on the intake.
+
+## Source
+
+- `AGENTS.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/harness.md`
+
+## Output
+
+```txt
+Result:
+Files:
+Risks:
+Next action:
+```

package/repo-template/.cursor/commands/sdd-tasks.md

@@ -0,0 +1,18 @@
+# /sdd-tasks
+
+Break technical plan into executable tasks.
+
+## Source
+
+- `AGENTS.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/harness.md`
+
+## Output
+
+```txt
+Result:
+Files:
+Risks:
+Next action:
+```

package/repo-template/.cursor/commands/sdd-validate.md

@@ -0,0 +1,18 @@
+# /sdd-validate
+
+Validate spec, tasks, tests/evals and implementation alignment.
+
+## Source
+
+- `AGENTS.md`
+- `docs/ai/spec-driven-development.md`
+- `docs/ai/harness.md`
+
+## Output
+
+```txt
+Result:
+Files:
+Risks:
+Next action:
+```

package/repo-template/.cursor/commands/spec-escalate.md

@@ -0,0 +1,9 @@
+# spec-escalate
+
+Review whether the current spec must be upgraded.
+
+Read:
+
+- `docs/skills/spec-escalation-review.md`
+
+Stop implementation if risk exceeds the current spec level.

package/repo-template/.cursor/commands/spec-intake.md

@@ -0,0 +1,10 @@
+# spec-intake
+
+Run spec intake and create the right spec template.
+
+Read:
+
+- `docs/skills/spec-intake.md`
+- `docs/skills/spec-complexity-classifier.md`
+
+Do not implement yet.

package/repo-template/.cursor/commands/spec-size.md

@@ -0,0 +1,21 @@
+# spec-size
+
+Classify the requested task as basic, standard or complex.
+
+Read:
+
+- `docs/ai/spec-sizing.md`
+- `docs/ai/spec-intake.md`
+- `docs/ai/spec-escalation.md`
+
+Return:
+
+```txt
+Spec level:
+Score:
+Hard triggers:
+Reason:
+Template:
+Validation:
+Approval:
+```

package/repo-template/.cursor/rules/adapter-parity.mdc

@@ -0,0 +1,20 @@
+---
+description: Adapter parity rule. Cursor is an IDE adapter, not the governance source.
+alwaysApply: true
+---
+
+# Adapter Parity Rule
+
+Cursor is IDE-primary only when the user chooses it for editing.
+
+Cursor is not governance-primary.
+
+Governance order:
+
+1. Current user instruction
+2. `AGENTS.md`
+3. `docs/ai/`
+4. `docs/skills/`
+5. `.cursor/`
+
+Read `docs/ai/adapter-parity.md` if adapter behavior is unclear.

package/repo-template/.cursor/rules/api.mdc

@@ -0,0 +1,12 @@
+---
+description: Reglas de API, contratos e integraciones.
+globs: "**/api/**,**/routes/**,**/services/**,**/server/**"
+alwaysApply: false
+---
+
+# API Rule
+
+Leer `docs/ai/api.md` antes de crear o modificar APIs.
+Validar request/response.
+Centralizar errores.
+Agregar integration test para endpoints nuevos.

package/repo-template/.cursor/rules/core.mdc

@@ -0,0 +1,21 @@
+---
+description: Regla central del proyecto. Aplica siempre.
+alwaysApply: true
+---
+
+# Core Agent Rule
+
+La fuente universal es `AGENTS.md`.
+La fuente técnica vive en `docs/ai/`.
+Los procedimientos viven en `docs/skills/`.
+
+Antes de generar código:
+
+1. Leer `AGENTS.md`.
+2. Leer el documento relevante en `docs/ai/`.
+3. Seguir el harness.
+4. Usar SDD para cambios significativos.
+5. Usar TDD para cambios de comportamiento.
+6. Usar evals para IA.
+
+No declarar listo si el pipeline falla.

package/repo-template/.cursor/rules/enforcement.mdc

@@ -0,0 +1,16 @@
+---
+description: Enforcement-first rules for agentic coding.
+alwaysApply: true
+---
+
+# Enforcement Rule
+
+Read:
+
+- `docs/ai/enforcement.md`
+- `docs/ai/quality-gates.md`
+- `docs/ai/trust-policy.md`
+- `docs/ai/eval-strategy.md`
+
+Do not treat prompts as enforcement.
+Critical changes require quality gates and human approval.

package/repo-template/.cursor/rules/evals.mdc

@@ -0,0 +1,10 @@
+---
+description: Reglas para cambios de IA, prompts, agentes y tool calling.
+globs: "**/prompts/**,**/agents/**,**/evals/**,**/*prompt*.*,**/*agent*.*"
+alwaysApply: false
+---
+
+# Evals Rule
+
+Cambios de IA requieren eval o justificación.
+Validar tool-call correctness, schema adherence, safety y prompt regression cuando aplique.

package/repo-template/.cursor/rules/git.mdc

@@ -0,0 +1,19 @@
+---
+description: Pipeline de git y commits.
+alwaysApply: true
+---
+
+# Git Rule
+
+Antes de commit:
+
+1. Diff review
+2. Secrets check
+3. Lint
+4. Format
+5. Typecheck
+6. Test
+7. Build
+8. Conventional commit
+
+Nunca commitear `.env`, secrets, logs temporales o cambios no relacionados.

package/repo-template/.cursor/rules/harness.mdc

@@ -0,0 +1,14 @@
+---
+description: Harness obligatorio para ejecución agéntica controlada.
+alwaysApply: true
+---
+
+# Harness Rule
+
+Todo cambio sigue:
+
+Requirement → Spec → Plan → Tests/Evals → Implementation → Validation → Review → Human Approval.
+
+Prohibido saltar tests en bugs.
+Prohibido implementar feature significativa sin spec.
+Prohibido cerrar tarea con pipeline fallido.

package/repo-template/.cursor/rules/loops.mdc

@@ -0,0 +1,17 @@
+---
+description: Loop Engineering rules for bounded agentic execution.
+alwaysApply: true
+---
+
+# Loop Rule
+
+No autonomous loop without boundaries.
+No repair loop without max attempts.
+No learning loop without evals.
+No production loop without observability.
+No critical loop without human approval.
+
+Read:
+- `docs/ai/loops.md`
+- `docs/ai/loop-policy.md`
+- `docs/ai/loop-observability.md`

package/repo-template/.cursor/rules/testing.mdc

@@ -0,0 +1,13 @@
+---
+description: Reglas de testing para cambios de comportamiento.
+globs: "**/*.test.*,**/*.spec.*,**/tests/**,**/test/**"
+alwaysApply: false
+---
+
+# Testing Rule
+
+Bug fix → regression test primero.
+Feature → unit/integration/E2E según impacto.
+Refactor → tests existentes verdes.
+
+Mocks solo para dependencias externas.

package/repo-template/.cursor/rules/ui.mdc

@@ -0,0 +1,13 @@
+---
+description: Reglas de UI y sistema de diseño.
+globs: "**/*.tsx,**/*.jsx,**/components/**,**/pages/**,**/app/**,**/views/**"
+alwaysApply: false
+---
+
+# UI Rule
+
+Leer `docs/ai/ui.md` antes de crear o modificar UI.
+No colores hardcodeados si existen tokens.
+Usar componentes base.
+Mobile-first.
+Accesibilidad obligatoria.

package/repo-template/.cursor/skills/api-design/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: api-design
+description: Endpoints, contratos e integraciones.
+---
+
+# API Design Skill
+
+Fuente de verdad:
+
+- `docs/ai/api.md`
+- `docs/skills/api-design.md`
+
+Usar cuando:
+
+Endpoints, contratos e integraciones.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/architecture/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: architecture
+description: Cambios estructurales, ADRs, límites de módulos.
+---
+
+# Architecture Skill
+
+Fuente de verdad:
+
+- `docs/ai/architecture.md`
+- `docs/skills/architecture.md`
+
+Usar cuando:
+
+Cambios estructurales, ADRs, límites de módulos.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/code-review/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: code-review
+description: Revisión técnica antes de merge o checkpoint.
+---
+
+# Code Review Skill
+
+Fuente de verdad:
+
+- `docs/ai/conventions.md`
+- `docs/skills/code-review.md`
+
+Usar cuando:
+
+Revisión técnica antes de merge o checkpoint.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/context-graph/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: context-graph
+description: Uso de Engram/Graphify, memoria y grafo contextual.
+---
+
+# Context Graph Skill
+
+Fuente de verdad:
+
+- `docs/ai/context-graph.md`
+- `docs/skills/context-graph.md`
+
+Usar cuando:
+
+Uso de Engram/Graphify, memoria y grafo contextual.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/debugging/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: debugging
+description: Bugs, errores, fallas y comportamiento inesperado.
+---
+
+# Debugging Skill
+
+Fuente de verdad:
+
+- `docs/ai/testing.md`
+- `docs/skills/debugging.md`
+
+Usar cuando:
+
+Bugs, errores, fallas y comportamiento inesperado.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/evals/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: evals
+description: Cambios de prompts, agentes, tool calling o RAG.
+---
+
+# AI Evals Skill
+
+Fuente de verdad:
+
+- `docs/ai/evals.md`
+- `docs/skills/evals.md`
+
+Usar cuando:
+
+Cambios de prompts, agentes, tool calling o RAG.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/git-workflow/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: git-workflow
+description: Commits, pipeline y checkpoint.
+---
+
+# Git Workflow Skill
+
+Fuente de verdad:
+
+- `docs/ai/git-workflow.md`
+- `docs/skills/git-workflow.md`
+
+Usar cuando:
+
+Commits, pipeline y checkpoint.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/testing/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: testing
+description: Tests, cobertura, regression tests, fixtures y mocks.
+---
+
+# Testing Skill
+
+Fuente de verdad:
+
+- `docs/ai/testing.md`
+- `docs/skills/testing.md`
+
+Usar cuando:
+
+Tests, cobertura, regression tests, fixtures y mocks.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.cursor/skills/ui-design/SKILL.md

@@ -0,0 +1,23 @@
+---
+name: ui-design
+description: Creación o modificación de UI.
+---
+
+# UI Design Skill
+
+Fuente de verdad:
+
+- `docs/ai/ui.md`
+- `docs/skills/ui-design.md`
+
+Usar cuando:
+
+Creación o modificación de UI.
+
+Procedimiento:
+
+1. Leer fuente de verdad.
+2. Identificar tipo de cambio.
+3. Aplicar SDD/TDD/evals según corresponda.
+4. Validar.
+5. Reportar.

package/repo-template/.gentle-ai/README.md

@@ -0,0 +1,37 @@
+# Gentle AI Adapter
+
+This folder documents how Gentle AI should interact with the universal harness.
+
+## Recommended commands
+
+```bash
+/sdd-init
+gentle-ai skill-registry refresh
+gentle-ai doctor
+```
+
+## Model routing
+
+Economic models such as DeepSeek can handle:
+
+- intake
+- task breakdown
+- scaffolding
+- simple tests
+- base documentation
+
+Escalate to a strong model or human for:
+
+- architecture
+- critical specs
+- security
+- acceptance review
+- complex debugging
+
+## Source of truth
+
+- `AGENTS.md`
+- `docs/ai/`
+- `docs/skills/`
+- `docs/specs/`
+- `evals/`

package/repo-template/.gentle-ai/loops/README.md

@@ -0,0 +1,27 @@
+# Gentle AI Loops
+
+These loop definitions complement Gentle AI SDD/TDD workflows inside OpenCode.
+
+## Recommended routing
+
+DeepSeek / economic model:
+
+- intake
+- task breakdown
+- scaffolding
+- simple tests
+- repair attempts 1-2
+- documentation base
+
+Strong model / human:
+
+- architecture
+- spec review
+- security
+- eval strategy
+- acceptance review
+- repeated loop failure
+
+## Rule
+
+No Gentle AI loop may run without boundaries from `docs/ai/loop-policy.md`.

package/repo-template/.gentle-ai/loops/sdd-tdd-repair.loop.md

@@ -0,0 +1,27 @@
+# Gentle AI Loop: SDD/TDD Repair
+
+## Input
+
+Spec + failing test.
+
+## Cycle
+
+1. Read spec.
+2. Read failing test.
+3. Implement minimal fix.
+4. Run test.
+5. Repair up to 3 attempts.
+6. Escalate if failing.
+
+## DeepSeek allowed
+
+- attempts 1-2
+- simple fixes
+- test scaffolding
+
+## Escalate
+
+- after repeated failure
+- architecture ambiguity
+- security/auth/data issue
+- changing spec needed

package/repo-template/.gentle-ai/model-routing.md

@@ -0,0 +1,25 @@
+# Gentle AI Model Routing
+
+## Economic model profile
+
+Use for:
+
+- intake
+- scaffolding
+- simple TDD
+- docs
+- mechanical refactors
+
+## Strong model profile
+
+Use for:
+
+- architecture
+- spec review
+- eval strategy
+- security review
+- acceptance review
+
+## Rule
+
+Never allow a low-cost model to approve high-impact architecture, security, auth, payment, data-loss or infra changes without review.

package/repo-template/.github/copilot-instructions.md

@@ -0,0 +1,37 @@
+# GitHub Copilot Instructions
+
+Read `AGENTS.md` first.
+
+Copilot is an adapter only. The universal core governs.
+
+Required guidance:
+
+- `docs/ai/governance.md`
+- `docs/ai/enforcement.md`
+- `docs/ai/quality-gates.md`
+- `docs/ai/trust-policy.md`
+- `docs/ai/eval-strategy.md`
+- `docs/ai/maintainability-gates.md`
+
+Rules:
+
+- Do not implement non-trivial changes without a spec.
+- Do not fix bugs without regression tests.
+- Do not change AI behavior without evals.
+- Do not approve high-impact changes.
+- Do not add dependencies without justification.
+- Do not ignore failing gates.
+
+## Spec sizing
+
+Before implementing non-trivial work, classify the task:
+
+```txt
+basic
+standard
+complex
+```
+
+Use `docs/ai/spec-sizing.md`.
+
+Do not treat high-risk work as a basic spec.