Organize Basic Memory
Reorganize notes into a coherent folder structure while preserving all knowledge graph relations.
Workflow
Step 1: Assess current structure
code
list_directory(dir_name="/", depth=3, project="<project>")
Map what exists: note counts per folder, naming patterns, any flat dumps.
Step 2: Propose folder structure
Based on note types and topics found, propose a structure. Common pattern:
code
project/
conversations/ # Session summaries, discussions
decisions/ # Decision records with rationale
discoveries/ # Troubleshooting, root cause findings
learnings/ # Techniques, insights, knowledge gained
planning/ # Roadmaps, action items, sprints
specs/ # Technical specifications
docs/ # Documentation, guides
archive/ # Outdated or superseded content
2025/ # Archived by year
Adapt to what the user actually has. Don't impose structure where it doesn't fit.
Present the proposed structure and get user approval before moving anything.
Step 3: Execute moves
For each note being moved:
code
move_note(
identifier="<exact title or permalink>",
destination_path="<folder>/<filename>.md",
project="<project>"
)
Key rules:
- •
move_notepreserves all relations automatically — no manual re-linking needed - •Destination path must include filename with
.mdextension - •Use the existing permalink as the filename (e.g.,
decisions/use-postgresql.md) - •Folders auto-created when specifying nested paths
Step 4: Archive stale content
For notes identified as outdated or superseded:
code
# Add deprecation notice
edit_note(
identifier="<old note>",
operation="prepend",
content="**ARCHIVED** - Superseded by [[<New Note>]]\n\n---\n\n",
project="<project>"
)
# Move to archive
move_note(
identifier="<old note>",
destination_path="archive/2025/<filename>.md",
project="<project>"
)
Step 5: Verify and report
code
list_directory(dir_name="/", depth=2, project="<project>")
Report to the user:
- •Notes moved: count
- •Notes archived: count
- •New folder structure (visual tree)
- •All relations preserved (automatic)