@fernando.zavaleta/ai-platform-core 0.0.7-beta

package/03-domains/03.1-qa/skills/normalize-ticket.md

@@ -0,0 +1,83 @@
+# Skill: Normalize Jira Ticket
+
+## Objective
+
+Clean and normalize Jira ticket data before QA analysis.
+
+Jira tickets often contain inconsistent structures.  
+This skill extracts and standardizes the data.
+
+---
+
+## Input
+
+File:
+
+05-output/agente-qa/[ISSUE_ID]/ticket.json
+
+---
+
+## Processing
+
+Extract and normalize the following fields:
+
+Ticket ID  
+Summary  
+Description  
+Acceptance Criteria  
+Components  
+Labels  
+Comments
+
+If acceptance criteria are not explicit:
+
+Infer them from:
+
+- Description
+- Comments
+- Business context
+
+---
+
+## Output
+
+Create:
+
+05-output/agente-qa/[ISSUE_ID]/[ISSUE_ID].normalized.md
+
+Structure:
+
+# Normalized Ticket
+
+## Ticket ID
+[ID]
+
+## Summary
+[text]
+
+## Description
+[text]
+
+## Acceptance Criteria
+- criteria 1
+- criteria 2
+- criteria 3
+
+## Technical Context
+
+Components:
+[list]
+
+Labels:
+[list]
+
+## Important Comments
+[list]
+
+---
+
+## Next Step
+
+Pass normalized data to:
+
+analyze-ticket-risk.md

package/03-domains/03.1-qa/skills/report-bug.md

@@ -0,0 +1,41 @@
+# Skill: Report Bug
+
+Objective:
+
+Generate a standardized bug report when a test case fails.
+
+---
+
+Input:
+
+Test case execution result
+
+Associated Jira ticket
+
+---
+
+Process:
+
+1 Identify the failing step
+2 Extract relevant information
+3 Generate a bug report using the template
+
+Template:
+
+templates/bug-report-template.md
+
+---
+
+Output:
+
+05-output/agente-qa/[JIRA_ID]/bugs/
+
+Example:
+
+[JIRA_ID]-bug-login-error.md
+
+---
+
+Validation:
+
+Ensure the issue type is BUG before submitting to Jira.

package/03-domains/03.1-qa/templates/bug-report-template.md

@@ -0,0 +1,62 @@
+# Bug Report
+
+## Summary
+
+Short description of the defect.
+
+---
+
+## Environment
+
+Example:
+
+QA / UAT / Production
+
+Browser / Device / API version
+
+---
+
+## Preconditions
+
+Conditions required before executing the test.
+
+---
+
+## Steps to Reproduce
+
+1 Step 1  
+2 Step 2  
+3 Step 3  
+
+---
+
+## Expected Result
+
+Describe the expected behavior.
+
+---
+
+## Actual Result
+
+Describe the actual behavior observed.
+
+---
+
+## Evidence
+
+Screenshots / logs / video.
+
+---
+
+## Severity
+
+Blocker  
+Critical  
+Major  
+Minor
+
+---
+
+## Linked Requirement
+
+Jira Ticket: [ID]

package/03-domains/03.1-qa/templates/test-case-template.md

@@ -0,0 +1,29 @@
+# Test Case
+
+## Title
+Validar <comportamiento>
+
+## Ticket
+Jira ID
+
+## Preconditions
+Condiciones necesarias antes de ejecutar la prueba.
+
+## Steps
+
+| Step | Expected Result |
+|-----|-----|
+| Step 1 | Result 1 |
+| Step 2 | Result 2 |
+| Step 3 | Result 3 |
+...
+| Step n | Result n |
+
+## Test Data
+Datos necesarios.
+
+## Priority
+High / Medium / Low
+
+## Type
+Functional / UI / Regression / Analytics

package/03-domains/03.1-qa/templates/test-plan-template.md

@@ -0,0 +1,86 @@
+# Test Plan
+
+## 1. Introduction
+Describe the purpose of the testing effort and the functionality being validated.
+
+## 2. Objectives
+Define the goals of testing.
+
+- Validate functionality
+- Detect defects
+- Ensure stability
+- Verify UX across platforms
+
+## 3. Testing Approach
+
+### Testing Types
+- Unit Testing
+- Manual Testing
+- Automated Testing
+- Smoke Testing
+- Regression Testing
+- Exploratory Testing
+
+## 4. In Scope
+List the user stories or epics included.
+
+| Priority | Type | Ticket | Description |
+|--------|--------|--------|--------|
+
+## 5. Summary
+Overview of test coverage.
+
+- Total test cases
+- Countries
+- Total testing days
+
+## 6. User Profiles
+List user types involved.
+
+Example:
+- New Consultant
+- Registered Consultant
+- Constant Consultant
+
+## 7. Test Environment
+List environments and devices.
+
+## 8. Tools
+
+TestRail  
+Jira  
+Automation Tools
+
+## 9. Data Management
+Test data required for execution.
+
+## 10. Test Schedule
+
+| Phase | Start | End |
+|------|------|------|
+
+## 11. Test Deliverables
+
+- Test Plan
+- Test Cases
+- Execution Report
+- Defect Report
+- Final Test Summary
+
+## 12. QA Resources
+
+| Role | Allocation | Notes |
+
+## 13. Entry Criteria
+
+Conditions required to start testing.
+
+## 14. Exit Criteria
+
+Conditions required to close testing.
+
+## 15. Defect Workflow
+
+Priority levels  
+Severity levels  
+Defect types

package/03-domains/03.3-l3/agent.md

@@ -0,0 +1,80 @@
+# Support L3 AI Agent Persona
+
+You are a Senior Support L3 Engineer AI Agent.
+
+Your mission is to diagnose production incidents by correlating Jira tickets with historical Root Cause Analysis (RCAs) from Confluence.
+
+All user-facing outputs must be in SPANISH.
+
+---
+
+# Support Workflow Pipeline
+
+Command:
+support.analyze [ISSUE_ID]
+
+---
+
+Step 1 — Fetch Incident Data
+Use: 02-mcp/02.1-jira/bridge-protocol.md
+Purpose: Retrieve technical details, logs, and symptoms from the Jira ticket.
+Store output as: 05-output/agente-l3/[ISSUE_ID]/ticket.json
+
+---
+
+Step 2 — Semantic RCA Search
+Use skill: 03-domains/03.3-support/skills/search-rca-confluence.md
+Mechanism: 
+- Extract "Error Keywords" and "Component" from ticket.json.
+- Execute Confluence MCP search specifically on: https://confluence.belcorp.biz/display/RSB/SBR+RCAs
+- Retrieve ONLY the top 3 most relevant RCAs to optimize token usage.
+
+---
+
+Step 3 — Incident Normalization
+Use skill: 03-domains/03.3-support/skills/normalize-incident.md
+Purpose: Clean logs and descriptions to focus on the technical failure.
+Output: [ISSUE_ID].normalized.md
+
+---
+
+Step 4 — Correlation & Diagnosis
+Use skill: 03-domains/03.3-support/skills/correlate-rca.md
+Purpose: Compare the current incident against the retrieved RCAs.
+Output: [ISSUE_ID].correlation-analysis.md
+
+---
+
+Step 5 — Final Diagnosis Report
+Use skill: 03-domains/03.3-support/skills/generate-diagnosis.md
+Purpose: Propose a Root Cause and Action Plan in Spanish.
+Output: [ISSUE_ID].diagnosis.md
+
+---
+
+# Storage & Artifacts
+
+All files must be stored in: 
+05-output/agente-l3/[ISSUE_ID]/
+
+Required Artifacts:
+- [ISSUE_ID].normalized.md
+- [ISSUE_ID].correlation-analysis.md
+- [ISSUE_ID].diagnosis.md
+
+---
+
+# Support Domain Context Rule
+- CONTEXT_LOCK: You are ONLY allowed to read and write within the '03-domains/03.3-support/' and '05-output/agente-l3/' directories.
+- DATA_ISOLATION: Any data, IDs, or context found in other domain folders (e.g., 03.1-qa) must be treated as non-existent for the current execution.
+- DYNAMIC_ID: The [ISSUE_ID] is strictly defined by the user command. Do not use historical IDs or examples from other domains.
+
+# Support L3 Agent - Scope Rules
+- **DOMAIN_ONLY**: Your operational scope is strictly limited to '03-domains/03.3-support/' and '05-output/agente-l3/'.
+- **COMMAND_PARAM_IS_TRUTH**: The [ISSUE_ID] is ONLY what follows the command 'support.analyze'. Never assume a ticket ID from existing files in '05-output/agente-qa/'.
+- **STRICT_OUTPUT**: All generated documentation MUST be saved in '05-output/agente-l3/[ISSUE_ID]/'.
+
+## Domain Isolation & Scoping Rules
+- **IDENTITY_LOCK**: You are a Support Agent. Do NOT use files, logic, or ticket IDs from `agente-qa` or `03.1-qa`.
+- **INPUT_PARAM**: The value provided after `support.analyze` is the ONLY valid [ISSUE_ID].
+- **DIRECTORY_RULE**: All outputs MUST be written to `05-output/agente-l3/[ISSUE_ID]/`. 

package/03-domains/03.3-l3/skills/analyze-incident-rca.md

@@ -0,0 +1,16 @@
+# Skill: Analyze Incident with RCA
+
+## Objective
+Diagnose a production incident by matching it against historical Root Cause Analysis (RCAs).
+
+## Workflow
+1. **Fetch Incident:** Retrieve Jira ticket data via `02-mcp/02.1-jira/bridge-protocol.md`.
+2. **Search RCA:** Use Confluence MCP to find the top 3 most similar RCAs.
+3. **Compare:** - Analyze "Symptoms" vs "Historical RCA Symptoms".
+   - Analyze "Component" vs "Affected RCA Component".
+4. **Output:** Generate a report in Spanish with the "Probable Root Cause" and "Suggested Fix".
+
+## Output Rules
+- Format: .md
+- Language: Spanish [cite: 1]
+- Location: 05-output/agente-l3/[INCIDENT_ID]/diagnosis.md

package/03-domains/03.3-l3/skills/correlate-rca.md

@@ -0,0 +1,16 @@
+# Skill: Correlation Analysis
+
+## Objective
+Analizar y comparar el incidente actual con el historial, asegurando que los hallazgos se mapeen correctamente al formato oficial.
+
+## Mandatory Reference
+- **Template Source:** `03-domains/03.3-l3/templates/rca-template.md`
+- **Constraint:** Use the structure defined in the referenced template to organize the correlation results.
+
+## Analysis Dimensions
+1. **Pattern Matching:** Comparar síntomas actuales con RCAs históricos.
+2. **Component Overlap:** Identificar si el componente fallido tiene antecedentes.
+3. **Template Alignment:** Clasificar data para las secciones de Root Cause y Prevention del template.
+
+## Output
+- `[ISSUE_ID].correlation-analysis.md`

package/03-domains/03.3-l3/skills/generate-diagnosis.md

@@ -0,0 +1,17 @@
+# Skill: Final Diagnosis Report
+
+## Objective
+Produce the final professional report for the L3 support team.
+
+## Requirements
+- Language: PROFESSIONAL SPANISH.
+- Format: Markdown (.md).
+
+## Structure
+1. **Resumen Ejecutivo:** Descripción de alto nivel.
+2. **Causa Raíz Probable:** Explicación técnica detallada.
+3. **Evidencia Histórica:** Resumen de por qué los RCAs encontrados son relevantes.
+4. **Plan de Acción:** Pasos sugeridos (Hotfix, Rollback, Configuration change).
+
+## Output
+- 05-output/agente-l3/[ISSUE_ID]/[ISSUE_ID].diagnosis.md

package/03-domains/03.3-l3/skills/normalize-incident.md

@@ -0,0 +1,15 @@
+# Skill: Normalize Incident Technical Data
+
+## Objective
+Clean and structure raw data from Jira (logs, descriptions, comments) to identify the core technical failure.
+
+## Input
+- 05-output/agente-l3/[ISSUE_ID]/ticket.json
+
+## Instructions
+1. **Sanitize Data:** Remove sensitive user information (PII) like names or phone numbers.
+2. **Log Extraction:** Identify stack traces, HTTP error codes (5xx, 4xx), and timeouts.
+3. **Symptom Identification:** Summarize what the user sees vs. what the system reports.
+
+## Output
+- [ISSUE_ID].normalized.md

package/03-domains/03.3-l3/skills/search-rca-confluence.md

@@ -0,0 +1,13 @@
+# Skill: Semantic RCA Search
+
+## Objective
+Find the most relevant historical RCAs without exceeding token limits.
+
+## Process
+1. **Keyword Generation:** Extract 3-5 technical keywords from the normalized incident (e.g., "MemoryLeak", "S3-Timeout").
+2. **Filtered Search:** - Execute search in: https://confluence.belcorp.biz/display/RSB/SBR+RCAs
+   - Limit results to the top 3 matches using the Confluence Bridge.
+3. **Data Retrieval:** Fetch only the "Root Cause" and "Resolution" sections of the matched RCAs.
+
+## Output
+- Intermediate JSON data (not stored in repo).

package/03-domains/03.3-l3/templates/rca-template.md

@@ -0,0 +1,21 @@
+# Root Cause Analysis (RCA) Standard Template
+
+## Mandatory Reference
+- **Official Source:** https://confluence.belcorp.biz/display/RSB/RCA+Template
+- **Constraint:** All findings must be structured to fit the sections defined below (Root Cause, Resolution, Impact, Prevention).
+
+---
+
+## Standard Structure for Agent Output
+
+### 1. Incident Overview
+(Summary of what happened)
+
+### 2. Root Cause Analysis
+(Technical explanation of why it happened)
+
+### 3. Impact Assessment
+(Business and system impact)
+
+### 4. Resolution & Prevention
+(How it was fixed and how to avoid it in the future)

package/03-domains/03.4-product/agent.md

@@ -0,0 +1,80 @@
+# Product Agent
+
+Role:
+AI Product Analyst
+
+Mission:
+
+Validate the quality and readiness of Jira stories before development begins.
+
+The agent evaluates whether the story meets the Definition of Ready.
+
+All outputs must be generated in SPANISH.
+
+---
+
+# Supported Commands
+
+Commands follow the pattern:
+
+product.[action] [JIRA_ID]
+
+Example:
+
+product.validate [JIRA_ID]
+
+---
+
+# Command: product.validate
+
+Purpose:
+
+Validate that a Jira ticket is properly defined and ready for development.
+
+Execution Flow:
+
+1 Fetch Jira ticket using MCP Bridge
+
+02-mcp/02.1-jira/bridge.md
+
+2 Execute validation skill
+
+skills/validate-story.md
+
+3 Generate readiness report
+
+---
+
+# Output
+
+Directory:
+
+05-output/agent-product/[JIRA_ID]/
+
+File generated:
+
+[JIRA_ID].story-validation.md
+
+---
+
+# Validation Criteria
+
+The agent must evaluate:
+
+Problem clarity  
+Business value  
+Acceptance criteria  
+Testability  
+Dependencies
+
+---
+
+# Examples
+
+Example command:
+
+product.validate [JIRA_ID]
+
+Expected result:
+
+Story validation report indicating readiness status.

package/03-domains/03.4-product/knowledge/story-quality.md

@@ -0,0 +1,16 @@
+# Story Quality Guidelines
+
+A good user story must follow the INVEST principles.
+
+Independent  
+Negotiable  
+Valuable  
+Estimable  
+Small  
+Testable
+
+Stories should clearly explain:
+
+Who needs the feature  
+What functionality is required  
+Why it is valuable