What is the Memory Bank?
The Memory Bank is a file-based storage system for all project artifacts. It maintains context across agent sessions and provides traceability between artifacts.Unlike traditional documentation that gets stale, the Memory Bank is actively used by agents. It’s the source of truth that agents read and write.
Why Memory Bank?
Context Engineering
Agents reload context from Memory Bank each session. No more lost knowledge.
Traceability
Every artifact links to its source. Inception and construction logs provide full traceability after completion.
Human Readable
All files are Markdown. Review, edit, and version control with Git.
AI Accessible
Structured format that agents can parse and update.
Structure
After project initialization:Artifact Types
Standards
Project-wide decisions that inform AI code generation:Intent Artifacts
Captured requirements and context:Unit Artifacts
Work breakdown within an intent:Bolt Artifacts
Implementation records:Traceability
Artifacts link to each other using references:Agent Interaction
Agents read and write to the Memory Bank:1
Context Loading
Agent reads relevant artifacts at session start
2
Work Execution
Agent generates new artifacts during work
3
Artifact Storage
Agent writes artifacts to Memory Bank
4
Reference Linking
Agent updates cross-references
Version Control
The Memory Bank is designed for Git:Schema
The Memory Bank follows a schema defined in.specsmd/aidlc/memory-bank.yaml:
Best Practices
Keep Artifacts Current
Keep Artifacts Current
Update artifacts when decisions change. Stale documentation is worse than no documentation.
Use Consistent Formatting
Use Consistent Formatting
Follow the templates. Consistent structure helps agents parse content.
Link Generously
Link Generously
Cross-reference related artifacts. Traceability prevents knowledge silos.
Version with Code
Version with Code
Commit Memory Bank changes with related code. They belong together.
Next Steps
Standards
Learn about project standards that guide AI generation