connect

Connect Notes

Connect a note to related notes in your vault, or discover connection opportunities across the vault.

Step 1: Load Configuration

Read ~/.claude/second-brain.md for vault name and path.

If missing:

Second brain not configured. Run /second-brain:setup first.

Use the symlink path ~/.claude/vaults/{name} to access the vault (e.g., ~/.claude/vaults/primary).

Load skill references:

• second-brain:obsidian for tool mechanics
• references/connecting.md for connection discovery algorithm

Step 2: Check Prerequisites

Follow references/connecting.md prerequisites:

which qmd

If qmd is not available:

qmd is not installed. Connection discovery requires qmd for semantic search.
Install via: bun add -g qmd
Then configure: qmd collection add ~/path/to/vault --name second-brain

Verify collection exists:

qmd collection list

Load collection name from ~/.claude/second-brain.md (qmd_collection setting). Default: second-brain.

Step 3: Determine Mode

If argument provided (note path): Go to Step 4 (Connect Specific Note) If no argument: Go to Step 7 (Connection Gardening)

---

Connect Specific Note

Step 4: Load Target Note

Resolve the path. Accept:

• Vault-relative path: Areas/gusto/202501121430 kafka-consumer-groups.md
• Absolute path: ~/Vaults/pickled-knowledge/.../file.md
• Just the filename: 202501121430 kafka-consumer-groups.md (search for it)

If just a filename, find it:

find "{vault}" -name "{filename}" -type f 2>/dev/null

Read the note content.

If note not found:

Note not found: {path}
Check the path and try again.

Step 5: Run Connection Discovery

Follow the full references/connecting.md algorithm:

1. Extract query terms from the note's title and body
2. Run qmd query "{terms}" -c {collection} -n 15 --json
3. Filter noise (self, already-linked, low-relevance, daily notes)
4. Enrich candidates (check for See Also, Related, TK sections)
5. Rank by connection value
6. Present candidates for multi-select