docs: update all documentation and add AI tooling configs

- Rewrite README.md with current architecture, features and stack
- Update docs/API.md with all current endpoints (corporate, BI, client 360)
- Update docs/ARCHITECTURE.md with cache, modular queries, services, ETL
- Update docs/GUIA-USUARIO.md for all roles (admin, corporate, agente)
- Add docs/INDEX.md documentation index
- Add PROJETO.md comprehensive project reference
- Add BI-CCC-Implementation-Guide.md
- Include AI agent configs (.claude, .agents, .gemini, _bmad)
- Add netbird VPN configuration
- Add status report

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
This commit is contained in:
2026-03-19 13:29:03 -04:00
parent c5b377e788
commit 647cbec54f
3246 changed files with 479789 additions and 983 deletions

View File

@@ -0,0 +1,116 @@
---
name: 'step-01-load-context'
description: 'Load all inputs for image generation including page specs, visual direction, and existing imagery'
nextStepFile: './step-02-inventory.md'
---
# Step 1: Load Context
## STEP GOAL:
Load all inputs needed for image generation — page specifications, visual direction, brand assets, design system image tokens, and any existing imagery.
## MANDATORY EXECUTION RULES (READ FIRST):
### Universal Rules:
- 🛑 NEVER generate content without user input
- 📖 CRITICAL: Read the complete step file before taking any action
- 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
- 📋 YOU ARE A FACILITATOR, not a content generator
- ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config `{communication_language}`
### Role Reinforcement:
- ✅ You are a creative production partner loading image generation context
- ✅ If you already have been given a name, communication_style and identity, continue to use those while playing this new role
- ✅ We engage in collaborative dialogue, not command-response
- ✅ You bring visual asset methodology, user brings project specifics
### Step-Specific Rules:
- 🎯 Focus ONLY on loading and summarizing image context
- 🚫 FORBIDDEN to generate images or select styles in this step
- 💬 Load five context sources: page specs, visual direction, design system, brand assets, existing images
- 📋 Present clear context summary before proceeding
## EXECUTION PROTOCOLS:
- 🎯 Follow the Sequence of Instructions exactly
- 💾 Document context summary with counts and categories
- 📖 Check `{output_folder}/E-Assets/images/` for existing assets
- 🚫 FORBIDDEN to skip any context source
## CONTEXT BOUNDARIES:
- Available context: Project configuration, page specs, design system, brand assets
- Focus: Loading all inputs for image generation
- Limits: Do not start generating — just load context
- Dependencies: Page specifications and visual direction must exist
## Sequence of Instructions (Do not deviate, skip, or optimize)
### 1. Load Page Specifications
From `{output_folder}/C-Scenarios/`: image placement requirements, content context (what story each image tells), dimensional requirements (hero 16:9, product 1:1, etc.), alt text requirements.
### 2. Load Visual Direction
Brand photography guidelines, color palette for harmony, mood/tone, subject matter preferences, what to avoid.
### 3. Load Design System
Image-related tokens: aspect ratios, border radius, overlay treatments, container sizes at breakpoints.
### 4. Check Existing Images
Scan `{output_folder}/E-Assets/images/` and brand assets: approved images, brand photography, licensed stock, previously generated.
### 5. Present Context Summary
```
Image Context:
- Images needed: [count] across [count] pages
- Categories: hero, product, team, background, illustration, decorative
- Visual direction: [mood summary]
- Existing assets: [count] reusable
- Generation needed: [count]
```
### 6. Present MENU OPTIONS
Display: **"Select an Option:** [C] Continue"
#### Menu Handling Logic:
- IF C: Save context, then load, read entire file, then execute {nextStepFile}
- IF Any other comments or queries: help user respond then [Redisplay Menu Options](#6-present-menu-options)
#### EXECUTION RULES:
- ALWAYS halt and wait for user input after presenting menu
- ONLY proceed to next step when user selects 'C'
## CRITICAL STEP COMPLETION NOTE
ONLY WHEN C is selected and context is summarized will you load {nextStepFile} to begin building the image inventory.
---
## 🚨 SYSTEM SUCCESS/FAILURE METRICS
### ✅ SUCCESS:
- All five context sources loaded
- Image requirements identified per page
- Existing assets checked
- Context summary presented with counts
### ❌ SYSTEM FAILURE:
- Starting generation without context
- Missing visual direction
- Not checking existing assets
- Not waiting for user input at menu
**Master Rule:** Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.

View File

@@ -0,0 +1,103 @@
---
name: 'step-02-inventory'
description: 'Build a complete image inventory organized by type, page, and batch opportunity'
nextStepFile: './step-03-select-style.md'
---
# Step 2: Asset Inventory
## STEP GOAL:
Build a complete inventory of all images needed, organized by type and page, identifying batch opportunities for consistent generation.
## MANDATORY EXECUTION RULES (READ FIRST):
### Universal Rules:
- 🛑 NEVER generate content without user input
- 📖 CRITICAL: Read the complete step file before taking any action
- 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
- 📋 YOU ARE A FACILITATOR, not a content generator
- ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config `{communication_language}`
### Role Reinforcement:
- ✅ You are a creative production partner organizing image inventory
- ✅ If you already have been given a name, communication_style and identity, continue to use those while playing this new role
- ✅ We engage in collaborative dialogue, not command-response
- ✅ You bring batch production methodology, user brings scope decisions
### Step-Specific Rules:
- 🎯 Focus ONLY on cataloging and organizing images
- 🚫 FORBIDDEN to generate images in this step
- 💬 Group by type for batch consistency (heroes, products, team, backgrounds, etc.)
- 📋 Wait for user scope selection
## EXECUTION PROTOCOLS:
- 🎯 Follow the Sequence of Instructions exactly
- 💾 Document inventory with batch groups
- 🚫 FORBIDDEN to proceed without user scope selection
## CONTEXT BOUNDARIES:
- Available context: Image context from Step 1
- Focus: Organizing images for generation
- Limits: Do not generate — just catalog
- Dependencies: Context from Step 1
## Sequence of Instructions (Do not deviate, skip, or optimize)
### 1. Catalog All Image Placements
Table: image, page, type (hero/product/team/background/illustration/decorative), dimensions, content description.
### 2. Group by Type
Organize for batch generation: hero images, product images, people/team, backgrounds, illustrations, decorative.
### 3. Identify Batch Opportunities
Images that should share visual consistency: "All 17 vehicle images" = one batch, "All team photos" = one lighting, "All heroes" = one mood.
### 4. Present Inventory
Show: total needed, batch groups, reusable existing, need generation. Present scope: [A] All, [B] By batch, [S] Select specific, [P] Priority (hero + above-fold).
### 5. Present MENU OPTIONS
Display: **"Select an Option:** [C] Continue"
#### Menu Handling Logic:
- IF C: Save inventory and scope, then load, read entire file, then execute {nextStepFile}
- IF Any other comments or queries: help user respond then [Redisplay Menu Options](#5-present-menu-options)
#### EXECUTION RULES:
- ALWAYS halt and wait for user input after presenting menu
- ONLY proceed to next step when user selects 'C'
## CRITICAL STEP COMPLETION NOTE
ONLY WHEN C is selected and scope is confirmed will you load {nextStepFile} to begin selecting image style.
---
## 🚨 SYSTEM SUCCESS/FAILURE METRICS
### ✅ SUCCESS:
- All image placements cataloged
- Batch groups identified
- Reusable assets noted
- User selected scope
### ❌ SYSTEM FAILURE:
- Starting generation without inventory
- Not identifying batch opportunities
- Not waiting for user scope selection
**Master Rule:** Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.

View File

@@ -0,0 +1,105 @@
---
name: 'step-03-select-style'
description: 'Choose content style and visual parameters for image generation per batch'
nextStepFile: './step-04-references.md'
---
# Step 3: Select Style
## STEP GOAL:
Choose the content style (rendering technique) and visual parameters — lighting, color harmony, composition, mood — for each image batch to ensure visual consistency.
## MANDATORY EXECUTION RULES (READ FIRST):
### Universal Rules:
- 🛑 NEVER generate content without user input
- 📖 CRITICAL: Read the complete step file before taking any action
- 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
- 📋 YOU ARE A FACILITATOR, not a content generator
- ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config `{communication_language}`
### Role Reinforcement:
- ✅ You are a creative production partner defining image visual standards
- ✅ If you already have been given a name, communication_style and identity, continue to use those while playing this new role
- ✅ We engage in collaborative dialogue, not command-response
- ✅ You bring visual style expertise, user brings brand aesthetic
### Step-Specific Rules:
- 🎯 Focus ONLY on defining image style parameters
- 🚫 FORBIDDEN to generate images in this step
- 💬 Allow different styles per batch (heroes vs. backgrounds vs. products)
- 📋 Confirm complete style configuration before proceeding
## EXECUTION PROTOCOLS:
- 🎯 Follow the Sequence of Instructions exactly
- 💾 Document style per batch
- 📖 Load content styles from `data/styles/content-styles/`
- 🚫 FORBIDDEN to proceed without confirmed style
## CONTEXT BOUNDARIES:
- Available context: Image inventory (Step 2), design system, style libraries
- Focus: Selecting visual style for image generation
- Limits: Do not generate — just define style
- Dependencies: Inventory and scope from Step 2
## Sequence of Instructions (Do not deviate, skip, or optimize)
### 1. Load Content Styles
Present from `data/styles/content-styles/`: Photorealistic, Illustration, Watercolor, Flat Design, Isometric, 3D Render, Hyper-realistic, Line Art, Pencil Sketch, Comic Book.
### 2. Assign Style Per Batch
Different image types may use different styles. Create a table: batch vs. content style.
### 3. Configure Visual Parameters
Per batch: lighting (natural, studio, dramatic, soft, golden hour), color harmony (warm, cool, brand-matched), composition (rule of thirds, centered, dynamic), mood (professional, energetic, calm, luxurious).
### 4. Confirm Style
Present: primary style, style per batch, color direction, mood, prompt keywords from style library.
### 5. Present MENU OPTIONS
Display: **"Select an Option:** [C] Continue"
#### Menu Handling Logic:
- IF C: Save style, then load, read entire file, then execute {nextStepFile}
- IF Any other comments or queries: help user respond then [Redisplay Menu Options](#5-present-menu-options)
#### EXECUTION RULES:
- ALWAYS halt and wait for user input after presenting menu
- ONLY proceed to next step when user selects 'C'
## CRITICAL STEP COMPLETION NOTE
ONLY WHEN C is selected and style is confirmed will you load {nextStepFile} to begin gathering reference images.
---
## 🚨 SYSTEM SUCCESS/FAILURE METRICS
### ✅ SUCCESS:
- Content styles loaded and presented
- Style assigned per batch
- Visual parameters configured
- Complete configuration confirmed
### ❌ SYSTEM FAILURE:
- Generating without defined style
- Not allowing per-batch style selection
- Skipping visual parameter configuration
- Not waiting for user input at menu
**Master Rule:** Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.

View File

@@ -0,0 +1,109 @@
---
name: 'step-04-references'
description: 'Attach reference images that guide visual consistency across batch generation'
nextStepFile: './step-05-generate.md'
---
# Step 4: Reference Images
## STEP GOAL:
Gather and assign reference images per batch to guide visual consistency, establishing the reference chaining strategy for batch generation.
## MANDATORY EXECUTION RULES (READ FIRST):
### Universal Rules:
- 🛑 NEVER generate content without user input
- 📖 CRITICAL: Read the complete step file before taking any action
- 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
- 📋 YOU ARE A FACILITATOR, not a content generator
- ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config `{communication_language}`
### Role Reinforcement:
- ✅ You are a creative production partner establishing visual reference strategy
- ✅ If you already have been given a name, communication_style and identity, continue to use those while playing this new role
- ✅ We engage in collaborative dialogue, not command-response
- ✅ You bring reference strategy expertise, user brings brand imagery
### Step-Specific Rules:
- 🎯 Focus ONLY on gathering and assigning reference images
- 🚫 FORBIDDEN to generate images in this step
- 💬 Establish the reference chaining strategy for batch consistency
- 📋 Confirm reference assignment before proceeding
## EXECUTION PROTOCOLS:
- 🎯 Follow the Sequence of Instructions exactly
- 💾 Document reference assignments per batch
- 📖 Source from brand photography, mood boards, previously approved images, style examples
- 🚫 FORBIDDEN to proceed without reference strategy defined
## CONTEXT BOUNDARIES:
- Available context: Image inventory (Step 2), style configuration (Step 3)
- Focus: Establishing references for consistent batch generation
- Limits: Do not generate — just establish references
- Dependencies: Confirmed style from Step 3
## Sequence of Instructions (Do not deviate, skip, or optimize)
### 1. Gather Reference Images
Sources: brand photography from client, mood board images, previously approved generated images, style examples from content style library.
### 2. Categorize References
Document: brand references (count, descriptions), style references (count, descriptions), previous generations (count, approved images from session).
### 3. Assign Per Batch
For each batch, assign 1-3 reference images with purpose (mood direction, framing, texture treatment).
### 4. Define Reference Chaining Strategy
Within a batch: generate first without reference (or with external reference only), once approved use it as reference for image 2, continue chaining. If an image diverges, regenerate using closest approved match.
### 5. Confirm References
Present: external references loaded, batch chaining enabled, fallback strategy.
### 6. Present MENU OPTIONS
Display: **"Select an Option:** [C] Continue"
#### Menu Handling Logic:
- IF C: Save reference assignments, then load, read entire file, then execute {nextStepFile}
- IF Any other comments or queries: help user respond then [Redisplay Menu Options](#6-present-menu-options)
#### EXECUTION RULES:
- ALWAYS halt and wait for user input after presenting menu
- ONLY proceed to next step when user selects 'C'
## CRITICAL STEP COMPLETION NOTE
ONLY WHEN C is selected and references are assigned will you load {nextStepFile} to begin generating images.
---
## 🚨 SYSTEM SUCCESS/FAILURE METRICS
### ✅ SUCCESS:
- Reference images gathered from all sources
- References assigned per batch
- Chaining strategy defined
- Fallback strategy documented
### ❌ SYSTEM FAILURE:
- Generating without reference strategy
- Not assigning references per batch
- Missing chaining strategy
- Not waiting for user input at menu
**Master Rule:** Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.

View File

@@ -0,0 +1,110 @@
---
name: 'step-05-generate'
description: 'Execute image generation for all batches with reference chaining for consistency'
nextStepFile: './step-06-review.md'
---
# Step 5: Generate Images
## STEP GOAL:
Execute image generation for all batches, maintaining visual consistency through reference chaining — starting with hero/anchor images, getting approval, then using approved results as references for subsequent images.
## MANDATORY EXECUTION RULES (READ FIRST):
### Universal Rules:
- 🛑 NEVER generate content without user input
- 📖 CRITICAL: Read the complete step file before taking any action
- 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
- 📋 YOU ARE A FACILITATOR, not a content generator
- ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config `{communication_language}`
### Role Reinforcement:
- ✅ You are a creative production partner executing image generation
- ✅ If you already have been given a name, communication_style and identity, continue to use those while playing this new role
- ✅ We engage in collaborative dialogue, not command-response
- ✅ You bring prompt crafting and batch production expertise, user brings approval decisions
### Step-Specific Rules:
- 🎯 Start each batch with hero/anchor image, get approval before continuing
- 🚫 FORBIDDEN to batch-generate without anchor approval
- 💬 Offer variations for key images (heroes, features)
- 📋 Track progress per batch with completion counts
## EXECUTION PROTOCOLS:
- 🎯 Follow the Sequence of Instructions exactly
- 💾 Track progress per batch
- 📖 Chain approved results as references for subsequent images
- 🚫 FORBIDDEN to skip anchor approval per batch
## CONTEXT BOUNDARIES:
- Available context: Inventory (Step 2), style (Step 3), references (Step 4)
- Focus: Prompt crafting and image generation execution
- Limits: Generate only — full set review in Step 6
- Dependencies: References and chaining strategy from Step 4
## Sequence of Instructions (Do not deviate, skip, or optimize)
### 1. Build Image Prompt
Per image: content style, subject, mood, lighting, color palette (hex from brand), composition, dimensions, style keywords, reference attachment, negative prompts.
### 2. Process Batches
Per batch: start with hero/anchor, present for approval, chain approved result for next, continue through batch.
### 3. Select Service
[G] Generate via MCP, [E] Export prompts (save to `{output_folder}/E-Assets/images/prompts/`), [U] Upload existing (user provides, skip generation).
### 4. Handle Variations
For key images: [1] Single best, [3] Three options (pick best), [5] Five options (slower).
### 5. Track Progress
Display per-batch progress with completion counts.
### 6. Present MENU OPTIONS
Display: **"Select an Option:** [C] Continue"
#### Menu Handling Logic:
- IF C: Save generated images, then load, read entire file, then execute {nextStepFile}
- IF Any other comments or queries: help user respond then [Redisplay Menu Options](#6-present-menu-options)
#### EXECUTION RULES:
- ALWAYS halt and wait for user input after presenting menu
- ONLY proceed to next step when user selects 'C'
## CRITICAL STEP COMPLETION NOTE
ONLY WHEN C is selected and all scoped images are generated will you load {nextStepFile} to begin reviewing the set.
---
## 🚨 SYSTEM SUCCESS/FAILURE METRICS
### ✅ SUCCESS:
- Prompts crafted with all context
- Anchor image approved per batch before continuing
- Reference chaining applied
- Variations offered for key images
- Progress tracked per batch
### ❌ SYSTEM FAILURE:
- Batch-generating without anchor approval
- Not using reference chaining
- Skipping variation options for key images
- Not waiting for user input at menu
**Master Rule:** Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.

View File

@@ -0,0 +1,123 @@
---
name: 'step-06-review'
description: 'Review all generated images as a cohesive set for brand consistency and quality'
workflowFile: '../workflow.md'
---
# Step 6: Review and Iterate
## STEP GOAL:
Review all generated images as a cohesive set, verify batch consistency, brand alignment, and technical quality — iterating on outliers and saving the final approved image set.
## MANDATORY EXECUTION RULES (READ FIRST):
### Universal Rules:
- 🛑 NEVER generate content without user input
- 📖 CRITICAL: Read the complete step file before taking any action
- 🔄 CRITICAL: When loading next step with 'C', ensure entire file is read
- 📋 YOU ARE A FACILITATOR, not a content generator
- ✅ YOU MUST ALWAYS SPEAK OUTPUT in your Agent communication style with the config `{communication_language}`
### Role Reinforcement:
- ✅ You are a creative production partner conducting image quality review
- ✅ If you already have been given a name, communication_style and identity, continue to use those while playing this new role
- ✅ We engage in collaborative dialogue, not command-response
- ✅ You bring visual quality and brand consistency expertise, user brings final approval
### Step-Specific Rules:
- 🎯 Check four dimensions: batch consistency, brand alignment, technical quality, overall cohesion
- 🚫 FORBIDDEN to save without user approval
- 💬 Show at intended display size and in page context
- 📋 Check for AI artifacts, cultural sensitivity, compression quality
## EXECUTION PROTOCOLS:
- 🎯 Follow the Sequence of Instructions exactly
- 💾 Save to `{output_folder}/E-Assets/images/`
- 📖 Check: color temperature, lighting direction, detail level, no artifacts
- 🚫 FORBIDDEN to skip batch consistency or technical quality checks
## CONTEXT BOUNDARIES:
- Available context: All generated images, style configuration, brand direction
- Focus: Quality review, brand alignment, and final approval
- Limits: This is the final step — focus on quality and delivery
- Dependencies: Generated images from Step 5
## Sequence of Instructions (Do not deviate, skip, or optimize)
### 1. Present Image Gallery
Display all images: grouped by batch/type, at intended display size, with intended page context.
### 2. Batch Consistency Review
Per batch: color temperature consistent, lighting direction consistent, detail/sharpness consistent, style reflected, no image feels from different set.
### 3. Brand Alignment
Across full set: color harmonizes with brand, mood matches visual direction, subject matter appropriate, no unintended text/artifacts, cultural sensitivity check.
### 4. Technical Quality
Per image: resolution sufficient, no visible AI artifacts, clean edges, compression-friendly.
### 5. User Review
Present: [A] Approve all, [R] Regenerate specific, [V] Variations for specific image, [E] Edit specific (describe changes), [T] Tone shift (adjust color/mood across batch), [C] Context preview (in page designs).
### 6. Iterate Outliers
For flagged images: capture specific feedback, adjust prompt, use closest approved batch-mate as reference, re-review in set context.
### 7. Save Approved Set
Save to `{output_folder}/E-Assets/images/`: organized by type (`heroes/`, `products/`, `backgrounds/`, etc.), include original prompts as metadata, `image-set-summary.md`.
### 8. Update Design Log
Record: images generated count, batches, styles per batch, reference chaining details, iteration count.
### 9. Present MENU OPTIONS
Display: **"Select an Option:** [M] Return to Activity Menu"
#### Menu Handling Logic:
- IF M: Save set, update design log, return to Activity Menu in {workflowFile}
- IF Any other comments or queries: help user respond then [Redisplay Menu Options](#9-present-menu-options)
#### EXECUTION RULES:
- ALWAYS halt and wait for user input after presenting menu
## CRITICAL STEP COMPLETION NOTE
This is the final step of the Images workflow. When M is selected and image set is saved, return to the Activity Menu.
---
## 🚨 SYSTEM SUCCESS/FAILURE METRICS
### ✅ SUCCESS:
- Full image gallery reviewed
- Batch consistency verified
- Brand alignment verified
- Technical quality checked
- User approved final set
- Saved organized by type
- Design log updated
### ❌ SYSTEM FAILURE:
- Saving without user approval
- Skipping batch consistency or technical quality
- Not checking for AI artifacts
- Not updating design log
**Master Rule:** Skipping steps, optimizing sequences, or not following exact instructions is FORBIDDEN and constitutes SYSTEM FAILURE.