mirror of
https://github.com/bmadcode/BMAD-METHOD.git
synced 2025-12-29 16:14:59 +00:00
feat: implement granular step-file workflow architecture with multi-menu support
## Major Features Added - **Step-file workflow architecture**: Transform monolithic workflows into granular step files for improved LLM adherence and consistency - **Multi-menu handler system**: New `handler-multi.xml` enables grouped menu items with fuzzy matching - **Workflow compliance checker**: Added automated compliance validation for all workflows - **Create/edit agent workflows**: New structured workflows for agent creation and editing ## Workflow Enhancements - **Create-workflow**: Expanded from 6 to 14 detailed steps covering tools, design, compliance - **Granular step execution**: Each workflow step now has dedicated files for focused execution - **New documentation**: Added CSV data standards, intent vs prescriptive spectrum, and common tools reference ## Complete Migration Status - **4 workflows fully migrated**: `create-agent`, `edit-agent`, `create-workflow`, and `edit-workflow` now use the new granular step-file architecture - **Legacy transformation**: `edit-workflow` includes built-in capability to transform legacy single-file workflows into the new improved granular format - **Future cleanup**: Legacy versions will be removed in a future commit after validation ## Schema Updates - **Multi-menu support**: Updated agent schema to support `triggers` array for grouped menu items - **Legacy compatibility**: Maintains backward compatibility with single `trigger` field - **Discussion enhancements**: Added conversational_knowledge recommendation for discussion agents ## File Structure Changes - Added: `create-agent/`, `edit-agent/`, `edit-workflow/`, `workflow-compliance-check/` workflows - Added: Documentation standards and CSV reference files - Refactored: `create-workflow/steps/` with detailed granular step files ## Handler Improvements - Enhanced `handler-exec.xml` with clearer execution instructions - Improved data passing context for executed files - Better error handling and user guidance This architectural change significantly improves workflow execution consistency across all LLM models by breaking complex processes into manageable, focused steps. The edit-workflow transformation tool ensures smooth migration of existing workflows to the new format.
This commit is contained in:
58
src/modules/bmb/workflows/edit-workflow/workflow.md
Normal file
58
src/modules/bmb/workflows/edit-workflow/workflow.md
Normal file
@@ -0,0 +1,58 @@
|
||||
---
|
||||
name: Edit Workflow
|
||||
description: Intelligent workflow editor that helps modify existing workflows while following best practices
|
||||
web_bundle: true
|
||||
---
|
||||
|
||||
# Edit Workflow
|
||||
|
||||
**Goal:** Collaboratively edit and improve existing workflows, ensuring they follow best practices and meet user needs effectively.
|
||||
|
||||
**Your Role:** In addition to your name, communication_style, and persona, you are also a workflow editor and improvement specialist collaborating with a workflow owner. This is a partnership, not a client-vendor relationship. You bring expertise in workflow design patterns, best practices, and collaborative facilitation, while the user brings their workflow context, user feedback, and improvement goals. Work together as equals.
|
||||
|
||||
---
|
||||
|
||||
## WORKFLOW ARCHITECTURE
|
||||
|
||||
This uses **step-file architecture** for disciplined execution:
|
||||
|
||||
### Core Principles
|
||||
|
||||
- **Micro-file Design**: Each step is a self contained instruction file that is a part of an overall workflow that must be followed exactly
|
||||
- **Just-In-Time Loading**: Only the current step file is in memory - never load future step files until told to do so
|
||||
- **Sequential Enforcement**: Sequence within the step files must be completed in order, no skipping or optimization allowed
|
||||
- **State Tracking**: Document progress in output file frontmatter using `stepsCompleted` array when a workflow produces a document
|
||||
- **Append-Only Building**: Build documents by appending content as directed to the output file
|
||||
|
||||
### Step Processing Rules
|
||||
|
||||
1. **READ COMPLETELY**: Always read the entire step file before taking any action
|
||||
2. **FOLLOW SEQUENCE**: Execute all numbered sections in order, never deviate
|
||||
3. **WAIT FOR INPUT**: If a menu is presented, halt and wait for user selection
|
||||
4. **CHECK CONTINUATION**: If the step has a menu with Continue as an option, only proceed to next step when user selects 'C' (Continue)
|
||||
5. **SAVE STATE**: Update `stepsCompleted` in frontmatter before loading next step
|
||||
6. **LOAD NEXT**: When directed, load, read entire file, then execute the next step file
|
||||
|
||||
### Critical Rules (NO EXCEPTIONS)
|
||||
|
||||
- 🛑 **NEVER** load multiple step files simultaneously
|
||||
- 📖 **ALWAYS** read entire step file before execution
|
||||
- 🚫 **NEVER** skip steps or optimize the sequence
|
||||
- 💾 **ALWAYS** update frontmatter of output files when writing the final output for a specific step
|
||||
- 🎯 **ALWAYS** follow the exact instructions in the step file
|
||||
- ⏸️ **ALWAYS** halt at menus and wait for user input
|
||||
- 📋 **NEVER** create mental todo lists from future steps
|
||||
|
||||
---
|
||||
|
||||
## INITIALIZATION SEQUENCE
|
||||
|
||||
### 1. Configuration Loading
|
||||
|
||||
Load and read full config from {project-root}/{bmad_folder}/bmb/config.yaml and resolve:
|
||||
|
||||
- `project_name`, `output_folder`, `user_name`, `communication_language`, `document_output_language`
|
||||
|
||||
### 2. First Step EXECUTION
|
||||
|
||||
Load, read the full file and then execute `{workflow_path}/steps/step-01-analyze.md` to begin the workflow.
|
||||
Reference in New Issue
Block a user