Responsible Vibe MCP

Appearance

Memory Systems ​

Responsible Vibe implements a three-layer context engineering approach that transforms AI from chaotic assistant to capable execution partner.

Overview on Layers of Context ​

💬

Conversation Memory

As outlined in how it works

⚙️

Process Memory

Phase-aware development plans

📚

Long-term Memory

Requirements, Architecture, Design

Layer 1: Conversation Memory ​

Systematic thinking and organized problem analysis

We outlined this in How it works

Layer 2: Process Memory ​

Phase-aware development plans and progress tracking

  • Current development phase and workflow state
  • What's been completed vs what's remaining
  • Decision history and reasoning

Layer 3: Long-term Memory ​

Requirements, Architecture, Design

  • Persistent project knowledge across sessions
  • Read on-demand
  • Created in early phases, permanently updated at the end of each feature

Process Memory: Development Plans ​

What It Is ​

Process memory is the development plan that steers the current conversation. Your AI actively maintains and updates this plan throughout the development process.

How it works:

  • Responsible Vibe creates a blank template with sections for each workflow phase
  • Your AI is fully in charge of maintaining what's in the plan
  • The AI updates tasks, marks completions, documents decisions
  • Used by whats_next() to determine current phase and next steps

The AI's Responsibility ​

markdown
## Requirements

### Tasks

- [ ] Understand user needs
- [ ] Document functional requirements
- [ ] Identify constraints

### Completed

- [x] Initial user interview
- [x] Core feature list defined

## Key Decisions

- Using JWT for authentication based on security requirements
- Terminal UI chosen for simplicity and cross-platform compatibility

The AI writes this content. Responsible Vibe only provides the structure.

How It Steers Conversation ​

When you call whats_next(), the tool:

  1. Reads the current development plan
  2. Analyzes what's complete vs incomplete
  3. Returns phase-specific instructions based on plan state
  4. Guides the AI on what to focus on next

This is active process memory - it directly controls the conversation flow.

Long-Term Memory: Project Documentation ​

What It Is ​

Long-term memory is structured project documentation that can be explicitly referenced when needed. Unlike process memory, this doesn't automatically influence conversations.

The .vibe/docs/ System ​

.vibe/
├── docs/
│   ├── architecture.md    # System design decisions
│   ├── requirements.md    # What you're building
│   └── design.md         # Implementation approach
└── development-plan-feature-auth.md  # Process memory (current)

Workflow Variable Substitution ​

Workflows can reference project documentation dynamically:

In Workflow Instructions:

"Review the system architecture documented in $ARCHITECTURE_DOC
and ensure your design addresses all requirements in $REQUIREMENTS_DOC."

At Runtime:

"Review the system architecture documented in /project/.vibe/docs/architecture.md
and ensure your design addresses all requirements in /project/.vibe/docs/requirements.md."

Explicit Reference System ​

Long-term memory requires explicit reference:

bash
# Reference in commits
git commit -m "implement user authentication

Based on security analysis in .vibe/docs/architecture.md,
implemented JWT with 24h expiry."

# Direct reference
"Check @.vibe/docs/architecture.md for the database schema decisions"

Key difference: Your AI must actively reference these documents - they don't automatically influence the conversation.

Setting Up Project Documentation ​

setup_project_docs Tool ​

Creates structured documentation for long-term reference:

Template Options:

  • arc42: Industry-standard architecture documentation
  • comprehensive: Detailed templates for all aspects
  • freestyle: Minimal structure, maximum flexibility
  • none: Placeholder that references plan file instead

File Linking:

bash
"Link the existing README.md as architecture documentation"
# Creates: .vibe/docs/architecture.md → README.md (symlink)

The Two Systems Working Together ​

Layer 2: Process Memory (Active) ​

  • Development plan maintained by AI
  • Automatically consulted by whats_next()
  • Steers current conversation and workflow phase
  • Updated continuously during development

Layer 3: Long-Term Memory (Passive) ​

  • Project documentation created by setup_project_docs
  • Referenced explicitly when needed
  • Survives across projects and conversations
  • Workflow variable substitution for consistent patterns

Layer 1 (Conversation Memory) is handled by your AI agent's natural conversation flow and systematic thinking patterns.

Real-World Example ​

bash
# AI maintains process memory (development plan)
## Implementation
### Tasks
- [x] Set up JWT authentication
- [ ] Add password hashing
- [ ] Implement session management

# You reference long-term memory when needed
"Look at @.vibe/docs/architecture.md to see how we decided to handle user sessions"

# Workflow automatically references long-term memory
"Ensure your implementation follows the security patterns in $ARCHITECTURE_DOC"

Why This Three-Layer Framework Matters ​

Layer 1 (Conversation Memory) provides systematic thinking and organized problem analysis within the current session.

Layer 2 (Process Memory) keeps your AI focused and on-track during active development with phase-aware guidance.

Layer 3 (Long-Term Memory) preserves architectural decisions and project knowledge that can be referenced when needed.

Together, these three layers provide the context AI needs to transform from chaotic assistant to capable execution partner - enabling both active guidance and reference documentation for serious software engineering.

Next: Advanced Engineering – Branch management and rule files integration