feat: voice notes with instant transcription and knowledge graph integration

- Voice memos now create notes immediately in knowledge/Voice Memos/<date>/
- Transcription shows directly in the note (Recording... → Transcribing... → transcript)
- Graph builder processes voice memos from knowledge directory
- Note creation agents now use workspace tools (writeFile, edit, grep, glob) instead of executeCommand
- Removes executeCommand dependency - no more permission prompts blocking knowledge updates

Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>

apps/x/apps/renderer/src/App.tsx

@@ -1461,6 +1461,29 @@ function AppContent({ auth }: { auth: ReturnType<typeof useRowboatAuth> }) {
     },
   }), [tree, selectedPath, workspaceRoot, collectDirPaths])
 
+  // Handler for when a voice note is created/updated
+  const handleVoiceNoteCreated = useCallback(async (notePath: string) => {
+    // Refresh the tree to show the new file/folder
+    const newTree = await loadDirectory()
+    setTree(newTree)
+
+    // Expand parent directories to show the file
+    const parts = notePath.split('/')
+    const parentPaths: string[] = []
+    for (let i = 1; i < parts.length; i++) {
+      parentPaths.push(parts.slice(0, i).join('/'))
+    }
+    setExpandedPaths(prev => {
+      const newSet = new Set(prev)
+      parentPaths.forEach(p => newSet.add(p))
+      return newSet
+    })
+
+    // Select the file to show it in the editor
+    setIsGraphOpen(false)
+    setSelectedPath(notePath)
+  }, [loadDirectory])
+
   const ensureWikiFile = useCallback(async (wikiPath: string) => {
     const resolvedPath = toKnowledgePath(wikiPath)
     if (!resolvedPath) return null
@@ -1707,6 +1730,7 @@ function AppContent({ auth }: { auth: ReturnType<typeof useRowboatAuth> }) {
               expandedPaths={expandedPaths}
               onSelectFile={toggleExpand}
               knowledgeActions={knowledgeActions}
+              onVoiceNoteCreated={handleVoiceNoteCreated}
               runs={runs}
               currentRunId={runId}
               tasksActions={{

apps/x/apps/renderer/src/components/sidebar-content.tsx

@@ -12,8 +12,10 @@ import {
   Folder,
   FolderPlus,
   MessageSquare,
+  Mic,
   Network,
   Pencil,
+  Square,
   SquarePen,
   Trash2,
 } from "lucide-react"
@@ -88,6 +90,7 @@ type SidebarContentPanelProps = {
   expandedPaths: Set<string>
   onSelectFile: (path: string, kind: "file" | "dir") => void
   knowledgeActions: KnowledgeActions
+  onVoiceNoteCreated?: (path: string) => void
   runs?: RunListItem[]
   currentRunId?: string | null
   tasksActions?: TasksActions
@@ -104,6 +107,7 @@ export function SidebarContentPanel({
   expandedPaths,
   onSelectFile,
   knowledgeActions,
+  onVoiceNoteCreated,
   runs = [],
   currentRunId,
   tasksActions,
@@ -126,6 +130,7 @@ export function SidebarContentPanel({
             expandedPaths={expandedPaths}
             onSelectFile={onSelectFile}
             actions={knowledgeActions}
+            onVoiceNoteCreated={onVoiceNoteCreated}
           />
         )}
         {activeSection === "tasks" && (
@@ -141,6 +146,227 @@ export function SidebarContentPanel({
   )
 }
 
+async function transcribeWithDeepgram(audioBlob: Blob): Promise<string | null> {
+  try {
+    const configResult = await window.ipc.invoke('workspace:readFile', {
+      path: 'config/deepgram.json',
+      encoding: 'utf8',
+    })
+    const { apiKey } = JSON.parse(configResult.data) as { apiKey: string }
+    if (!apiKey) throw new Error('No apiKey in deepgram.json')
+
+    const response = await fetch(
+      'https://api.deepgram.com/v1/listen?model=nova-2&smart_format=true',
+      {
+        method: 'POST',
+        headers: {
+          Authorization: `Token ${apiKey}`,
+          'Content-Type': audioBlob.type,
+        },
+        body: audioBlob,
+      },
+    )
+
+    if (!response.ok) throw new Error(`Deepgram API error: ${response.status}`)
+    const result = await response.json()
+    return result.results?.channels?.[0]?.alternatives?.[0]?.transcript ?? null
+  } catch (err) {
+    console.error('Deepgram transcription failed:', err)
+    return null
+  }
+}
+
+// Voice Note Recording Button
+function VoiceNoteButton({ onNoteCreated }: { onNoteCreated?: (path: string) => void }) {
+  const [isRecording, setIsRecording] = React.useState(false)
+  const mediaRecorderRef = React.useRef<MediaRecorder | null>(null)
+  const chunksRef = React.useRef<Blob[]>([])
+  const notePathRef = React.useRef<string | null>(null)
+  const timestampRef = React.useRef<string | null>(null)
+  const relativePathRef = React.useRef<string | null>(null)
+
+  const startRecording = async () => {
+    try {
+      // Generate timestamp and paths immediately
+      const now = new Date()
+      const timestamp = now.toISOString().replace(/[:.]/g, '-')
+      const dateStr = now.toISOString().split('T')[0] // YYYY-MM-DD
+      const noteName = `voice-memo-${timestamp}`
+      const notePath = `knowledge/Voice Memos/${dateStr}/${noteName}.md`
+
+      timestampRef.current = timestamp
+      notePathRef.current = notePath
+      // Relative path for linking (from knowledge/ root, without .md extension)
+      const relativePath = `Voice Memos/${dateStr}/${noteName}`
+      relativePathRef.current = relativePath
+
+      // Create the note immediately with a "Recording..." placeholder
+      await window.ipc.invoke('workspace:mkdir', {
+        path: `knowledge/Voice Memos/${dateStr}`,
+        recursive: true,
+      })
+
+      const initialContent = `# Voice Memo
+
+**Type:** voice memo
+**Recorded:** ${now.toLocaleString()}
+**Path:** ${relativePath}
+
+## Transcript
+
+*Recording in progress...*
+`
+      await window.ipc.invoke('workspace:writeFile', {
+        path: notePath,
+        data: initialContent,
+        opts: { encoding: 'utf8' },
+      })
+
+      // Select the note so the user can see it
+      onNoteCreated?.(notePath)
+
+      // Start actual recording
+      const stream = await navigator.mediaDevices.getUserMedia({ audio: true })
+      const mimeType = MediaRecorder.isTypeSupported('audio/mp4')
+        ? 'audio/mp4'
+        : 'audio/webm'
+      const recorder = new MediaRecorder(stream, { mimeType })
+      chunksRef.current = []
+
+      recorder.ondataavailable = (e) => {
+        if (e.data.size > 0) chunksRef.current.push(e.data)
+      }
+
+      recorder.onstop = async () => {
+        stream.getTracks().forEach((t) => t.stop())
+        const blob = new Blob(chunksRef.current, { type: mimeType })
+        const ext = mimeType === 'audio/mp4' ? 'm4a' : 'webm'
+        const audioFilename = `voice-memo-${timestampRef.current}.${ext}`
+
+        // Save audio file to voice_memos folder (for backup/reference)
+        try {
+          await window.ipc.invoke('workspace:mkdir', {
+            path: 'voice_memos',
+            recursive: true,
+          })
+
+          const arrayBuffer = await blob.arrayBuffer()
+          const base64 = btoa(
+            new Uint8Array(arrayBuffer).reduce(
+              (data, byte) => data + String.fromCharCode(byte),
+              '',
+            ),
+          )
+
+          await window.ipc.invoke('workspace:writeFile', {
+            path: `voice_memos/${audioFilename}`,
+            data: base64,
+            opts: { encoding: 'base64' },
+          })
+        } catch {
+          console.error('Failed to save audio file')
+        }
+
+        // Update note to show transcribing status
+        const currentNotePath = notePathRef.current
+        const currentRelativePath = relativePathRef.current
+        if (currentNotePath && currentRelativePath) {
+          const transcribingContent = `# Voice Memo
+
+**Type:** voice memo
+**Recorded:** ${new Date().toLocaleString()}
+**Path:** ${currentRelativePath}
+
+## Transcript
+
+*Transcribing...*
+`
+          await window.ipc.invoke('workspace:writeFile', {
+            path: currentNotePath,
+            data: transcribingContent,
+            opts: { encoding: 'utf8' },
+          })
+        }
+
+        // Transcribe and update the note with the transcript
+        const transcript = await transcribeWithDeepgram(blob)
+        if (currentNotePath && currentRelativePath) {
+          const finalContent = transcript
+            ? `# Voice Memo
+
+**Type:** voice memo
+**Recorded:** ${new Date().toLocaleString()}
+**Path:** ${currentRelativePath}
+
+## Transcript
+
+${transcript}
+`
+            : `# Voice Memo
+
+**Type:** voice memo
+**Recorded:** ${new Date().toLocaleString()}
+**Path:** ${currentRelativePath}
+
+## Transcript
+
+*Transcription failed. Please try again.*
+`
+          await window.ipc.invoke('workspace:writeFile', {
+            path: currentNotePath,
+            data: finalContent,
+            opts: { encoding: 'utf8' },
+          })
+
+          // Re-select to trigger refresh
+          onNoteCreated?.(currentNotePath)
+
+          if (transcript) {
+            toast('Voice note transcribed', 'success')
+          } else {
+            toast('Transcription failed', 'error')
+          }
+        }
+      }
+
+      recorder.start()
+      mediaRecorderRef.current = recorder
+      setIsRecording(true)
+      toast('Recording started', 'success')
+    } catch {
+      toast('Could not access microphone', 'error')
+    }
+  }
+
+  const stopRecording = () => {
+    if (mediaRecorderRef.current && mediaRecorderRef.current.state !== 'inactive') {
+      mediaRecorderRef.current.stop()
+    }
+    mediaRecorderRef.current = null
+    setIsRecording(false)
+  }
+
+  return (
+    <Tooltip>
+      <TooltipTrigger asChild>
+        <button
+          onClick={isRecording ? stopRecording : startRecording}
+          className="text-sidebar-foreground/70 hover:text-sidebar-foreground hover:bg-sidebar-accent rounded p-1.5 transition-colors"
+        >
+          {isRecording ? (
+            <Square className="size-4 fill-red-500 text-red-500 animate-pulse" />
+          ) : (
+            <Mic className="size-4" />
+          )}
+        </button>
+      </TooltipTrigger>
+      <TooltipContent side="bottom">
+        {isRecording ? 'Stop Recording' : 'New Voice Note'}
+      </TooltipContent>
+    </Tooltip>
+  )
+}
+
 // Knowledge Section
 function KnowledgeSection({
   tree,
@@ -148,15 +374,17 @@ function KnowledgeSection({
   expandedPaths,
   onSelectFile,
   actions,
+  onVoiceNoteCreated,
 }: {
   tree: TreeNode[]
   selectedPath: string | null
   expandedPaths: Set<string>
   onSelectFile: (path: string, kind: "file" | "dir") => void
   actions: KnowledgeActions
+  onVoiceNoteCreated?: (path: string) => void
 }) {
   const isExpanded = expandedPaths.size > 0
-  
+
   const quickActions = [
     { icon: FilePlus, label: "New Note", action: () => actions.createNote() },
     { icon: FolderPlus, label: "New Folder", action: () => actions.createFolder() },
@@ -181,6 +409,7 @@ function KnowledgeSection({
                 <TooltipContent side="bottom">{action.label}</TooltipContent>
               </Tooltip>
             ))}
+            <VoiceNoteButton onNoteCreated={onVoiceNoteCreated} />
             <Tooltip>
               <TooltipTrigger asChild>
                 <button

apps/x/packages/core/src/knowledge/build_graph.ts

@@ -29,7 +29,89 @@ const SOURCE_FOLDERS = [
     'fireflies_transcripts',
     'granola_notes',
 ];
-const MAX_CONCURRENT_BATCHES = 1; // Process only 1 batch at a time to avoid overwhelming the agent
+
+// Voice memos are now created directly in knowledge/Voice Memos/<date>/
+const VOICE_MEMOS_KNOWLEDGE_DIR = path.join(NOTES_OUTPUT_DIR, 'Voice Memos');
+
+/**
+ * Get unprocessed voice memo files from knowledge/Voice Memos/
+ * Voice memos are created directly in this directory by the UI.
+ * Returns paths to files that need entity extraction.
+ */
+function getUnprocessedVoiceMemos(state: GraphState): string[] {
+    console.log(`[GraphBuilder] Checking directory: ${VOICE_MEMOS_KNOWLEDGE_DIR}`);
+
+    if (!fs.existsSync(VOICE_MEMOS_KNOWLEDGE_DIR)) {
+        console.log(`[GraphBuilder] Directory does not exist`);
+        return [];
+    }
+
+    const unprocessedFiles: string[] = [];
+
+    // Scan date folders (e.g., 2026-02-03)
+    const dateFolders = fs.readdirSync(VOICE_MEMOS_KNOWLEDGE_DIR);
+    console.log(`[GraphBuilder] Found ${dateFolders.length} date folders: ${dateFolders.join(', ')}`);
+
+    for (const dateFolder of dateFolders) {
+        const dateFolderPath = path.join(VOICE_MEMOS_KNOWLEDGE_DIR, dateFolder);
+
+        // Skip if not a directory
+        try {
+            if (!fs.statSync(dateFolderPath).isDirectory()) {
+                continue;
+            }
+        } catch (err) {
+            console.log(`[GraphBuilder] Error checking ${dateFolderPath}:`, err);
+            continue;
+        }
+
+        // Scan markdown files in this date folder
+        const files = fs.readdirSync(dateFolderPath);
+        console.log(`[GraphBuilder] Found ${files.length} files in ${dateFolder}: ${files.join(', ')}`);
+
+        for (const file of files) {
+            // Only process voice memo markdown files
+            if (!file.endsWith('.md') || !file.startsWith('voice-memo-')) {
+                console.log(`[GraphBuilder] Skipping ${file} - not a voice memo file`);
+                continue;
+            }
+
+            const filePath = path.join(dateFolderPath, file);
+
+            // Skip if already processed
+            if (state.processedFiles[filePath]) {
+                console.log(`[GraphBuilder] Skipping ${file} - already processed`);
+                continue;
+            }
+
+            // Check if the file has actual content (not still recording/transcribing)
+            try {
+                const content = fs.readFileSync(filePath, 'utf-8');
+                // Skip files that are still recording or transcribing
+                if (content.includes('*Recording in progress...*')) {
+                    console.log(`[GraphBuilder] Skipping ${file} - still recording`);
+                    continue;
+                }
+                if (content.includes('*Transcribing...*')) {
+                    console.log(`[GraphBuilder] Skipping ${file} - still transcribing`);
+                    continue;
+                }
+                if (content.includes('*Transcription failed')) {
+                    console.log(`[GraphBuilder] Skipping ${file} - transcription failed`);
+                    continue;
+                }
+                console.log(`[GraphBuilder] Found unprocessed voice memo: ${file}`);
+                unprocessedFiles.push(filePath);
+            } catch (err) {
+                console.log(`[GraphBuilder] Error reading ${file}:`, err);
+                continue;
+            }
+        }
+    }
+
+    console.log(`[GraphBuilder] Total unprocessed files: ${unprocessedFiles.length}`);
+    return unprocessedFiles;
+}
 
 /**
  * Read content for specific files
@@ -186,6 +268,69 @@ export async function buildGraph(sourceDir: string): Promise<void> {
     console.log(`Knowledge graph build complete. Processed ${processedFiles.length} files.`);
 }
 
+/**
+ * Process voice memos from knowledge/Voice Memos/ and run entity extraction on them
+ * Voice memos are now created directly in the knowledge directory by the UI.
+ */
+async function processVoiceMemosForKnowledge(): Promise<boolean> {
+    console.log(`[GraphBuilder] Starting voice memo processing...`);
+    const state = loadState();
+
+    // Get unprocessed voice memos from knowledge/Voice Memos/
+    const unprocessedFiles = getUnprocessedVoiceMemos(state);
+
+    if (unprocessedFiles.length === 0) {
+        console.log(`[GraphBuilder] No unprocessed voice memos found`);
+        return false;
+    }
+
+    console.log(`[GraphBuilder] Processing ${unprocessedFiles.length} voice memo transcripts for entity extraction...`);
+    console.log(`[GraphBuilder] Files to process: ${unprocessedFiles.map(f => path.basename(f)).join(', ')}`);
+
+    // Read the files
+    const contentFiles = await readFileContents(unprocessedFiles);
+
+    if (contentFiles.length === 0) {
+        return false;
+    }
+
+    // Process in batches like other sources
+    const BATCH_SIZE = 10;
+    const totalBatches = Math.ceil(contentFiles.length / BATCH_SIZE);
+
+    for (let i = 0; i < contentFiles.length; i += BATCH_SIZE) {
+        const batch = contentFiles.slice(i, i + BATCH_SIZE);
+        const batchNumber = Math.floor(i / BATCH_SIZE) + 1;
+
+        try {
+            // Build knowledge index
+            console.log(`[GraphBuilder] Building knowledge index for batch ${batchNumber}...`);
+            const index = buildKnowledgeIndex();
+            const indexForPrompt = formatIndexForPrompt(index);
+
+            console.log(`[GraphBuilder] Processing batch ${batchNumber}/${totalBatches} (${batch.length} files)...`);
+            await createNotesFromBatch(batch, batchNumber, indexForPrompt);
+            console.log(`[GraphBuilder] Batch ${batchNumber}/${totalBatches} complete`);
+
+            // Mark files as processed
+            for (const file of batch) {
+                markFileAsProcessed(file.path, state);
+            }
+
+            // Save state after each batch
+            saveState(state);
+        } catch (error) {
+            console.error(`[GraphBuilder] Error processing batch ${batchNumber}:`, error);
+        }
+    }
+
+    // Update last build time
+    state.lastBuildTime = new Date().toISOString();
+    saveState(state);
+
+    return true;
+}
+
 /**
  * Process all configured source directories
  */
@@ -197,6 +342,16 @@ async function processAllSources(): Promise<void> {
 
     let anyFilesProcessed = false;
 
+    // Process voice memos first (they get moved to knowledge/)
+    try {
+        const voiceMemosProcessed = await processVoiceMemosForKnowledge();
+        if (voiceMemosProcessed) {
+            anyFilesProcessed = true;
+        }
+    } catch (error) {
+        console.error('[GraphBuilder] Error processing voice memos:', error);
+    }
+
     for (const folder of SOURCE_FOLDERS) {
         const sourceDir = path.join(WorkDir, folder);
 
@@ -234,7 +389,7 @@ async function processAllSources(): Promise<void> {
  */
 export async function init() {
     console.log('[GraphBuilder] Starting Knowledge Graph Builder Service...');
-    console.log(`[GraphBuilder] Monitoring folders: ${SOURCE_FOLDERS.join(', ')}`);
+    console.log(`[GraphBuilder] Monitoring folders: ${SOURCE_FOLDERS.join(', ')}, knowledge/Voice Memos`);
     console.log(`[GraphBuilder] Will check for new content every ${SYNC_INTERVAL_MS / 1000} seconds`);
 
     // Initial run

apps/x/packages/core/src/knowledge/note_creation_high.ts

@@ -7,19 +7,25 @@ tools:
   workspace-readFile:
     type: builtin
     name: workspace-readFile
+  workspace-edit:
+    type: builtin
+    name: workspace-edit
   workspace-readdir:
     type: builtin
     name: workspace-readdir
   workspace-mkdir:
     type: builtin
     name: workspace-mkdir
-  executeCommand:
+  workspace-grep:
     type: builtin
-    name: executeCommand
+    name: workspace-grep
+  workspace-glob:
+    type: builtin
+    name: workspace-glob
 ---
 # Task
 
-You are a memory agent. Given a single source file (email or meeting transcript), you will:
+You are a memory agent. Given a single source file (email, meeting transcript, or voice memo), you will:
 
 1. **Determine source type (meeting or email)**
 2. **Evaluate if the source is worth processing**
@@ -31,7 +37,7 @@ You are a memory agent. Given a single source file (email or meeting transcript)
 8. Create new notes (meetings only) or update existing notes
 9. **Apply state changes to existing notes**
 
-The core rule: **Meetings create notes. Emails enrich them.**
+The core rule: **Meetings and voice memos create notes. Emails enrich them.**
 
 You have full read access to the existing knowledge directory. Use this extensively to:
 - Find existing notes for people, organizations, projects mentioned
@@ -70,20 +76,51 @@ When you need to:
 
 # Tools Available
 
-You have access to \`executeCommand\` to run shell commands:
+You have access to these tools:
+
+**For reading files:**
 \`\`\`
-executeCommand("ls {path}")                     # List directory contents
+workspace-readFile({ path: "knowledge/People/Sarah Chen.md" })
-executeCommand("cat {path}")                    # Read file contents
-executeCommand("head -50 {path}")               # Read first 50 lines
-executeCommand("write {path} {content}")        # Create or overwrite file
 \`\`\`
 
-**Important:** Use shell escaping for paths with spaces:
+**For creating NEW files:**
 \`\`\`
-executeCommand("cat 'knowledge_folder/People/Sarah Chen.md'")
+workspace-writeFile({ path: "knowledge/People/Sarah Chen.md", data: "# Sarah Chen\\n\\n..." })
 \`\`\`
 
-**NOTE:** Do NOT use grep to search for entities. Use the provided knowledge_index instead.
+**For editing EXISTING files (preferred for updates):**
+\`\`\`
+workspace-edit({
+  path: "knowledge/People/Sarah Chen.md",
+  oldString: "## Activity\\n",
+  newString: "## Activity\\n- **2026-02-03** (meeting): New activity entry\\n"
+})
+\`\`\`
+
+**For listing directories:**
+\`\`\`
+workspace-readdir({ path: "knowledge/People" })
+\`\`\`
+
+**For creating directories:**
+\`\`\`
+workspace-mkdir({ path: "knowledge/Projects", recursive: true })
+\`\`\`
+
+**For searching files:**
+\`\`\`
+workspace-grep({ pattern: "Acme Corp", searchPath: "knowledge", fileGlob: "*.md" })
+\`\`\`
+
+**For finding files by pattern:**
+\`\`\`
+workspace-glob({ pattern: "**/*.md", cwd: "knowledge/People" })
+\`\`\`
+
+**IMPORTANT:**
+- Use \`workspace-edit\` for updating existing notes (adding activity, updating fields)
+- Use \`workspace-writeFile\` only for creating new notes
+- Prefer the knowledge_index for entity resolution (it's faster than grep)
 
 # Output
 
@@ -113,7 +150,7 @@ Either:
 
 Read the source file and determine if it's a meeting or email.
 \`\`\`
-executeCommand("cat '{source_file}'")
+workspace-readFile({ path: "{source_file}" })
 \`\`\`
 
 **Meeting indicators:**
@@ -126,9 +163,15 @@ executeCommand("cat '{source_file}'")
 - Has \`Subject:\` field
 - Email signature
 
+**Voice memo indicators:**
+- Has \`**Type:** voice memo\` field
+- Has \`**Path:**\` field with path like \`Voice Memos/YYYY-MM-DD/...\`
+- Has \`## Transcript\` section
+
 **Set processing mode:**
 - \`source_type = "meeting"\` → Can create new notes
 - \`source_type = "email"\` → Can only update existing notes
+- \`source_type = "voice_memo"\` → Can create new notes (treat like meetings)
 
 ---
 
@@ -301,9 +344,9 @@ If someone only appears in your memory as "CC'd on outreach emails from [Sender]
 ## Email-Specific Filtering
 
 For emails, check if sender/recipients have existing notes:
-\`\`\`bash
+\`\`\`
-executeCommand("grep -r -i -l '{sender email}' '{knowledge_folder}/'")
+workspace-grep({ pattern: "{sender email}", searchPath: "{knowledge_folder}" })
-executeCommand("grep -r -i -l '{sender name}' '{knowledge_folder}/People/'")
+workspace-grep({ pattern: "{sender name}", searchPath: "{knowledge_folder}/People" })
 \`\`\`
 
 **If no existing note found:**
@@ -343,7 +386,7 @@ If processing, continue to Step 2.
 
 # Step 2: Read and Parse Source File
 \`\`\`
-executeCommand("cat '{source_file}'")
+workspace-readFile({ path: "{source_file}" })
 \`\`\`
 
 Extract metadata:
@@ -440,7 +483,7 @@ From index, find matches for:
 
 Only read the full note content when you need details not in the index (e.g., activity logs, open items):
 \`\`\`bash
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-readFile({ path: "{knowledge_folder}/People/Sarah Chen.md" })
 \`\`\`
 
 **Why read these notes:**
@@ -524,27 +567,27 @@ Resolution Map:
 When multiple candidates match a variant, disambiguate:
 
 **By organization (strongest signal):**
-\`\`\`bash
+\`\`\`
 # "David" could be David Kim or David Chen
-executeCommand("grep -i 'Acme' '{knowledge_folder}/People/David Kim.md'")
+workspace-grep({ pattern: "Acme", searchPath: "{knowledge_folder}/People/David Kim.md" })
 # Output: **Organization:** [[Acme Corp]]
 
-executeCommand("grep -i 'Acme' '{knowledge_folder}/People/David Chen.md'")
+workspace-grep({ pattern: "Acme", searchPath: "{knowledge_folder}/People/David Chen.md" })
 # Output: **Organization:** [[Other Corp]]
 
 # Source is from Acme context → "David" = "David Kim"
 \`\`\`
 
 **By email (definitive):**
-\`\`\`bash
+\`\`\`
-executeCommand("grep -i 'david@acme.com' '{knowledge_folder}/People/David Kim.md'")
+workspace-grep({ pattern: "david@acme.com", searchPath: "{knowledge_folder}/People/David Kim.md" })
 # Exact email match is definitive
 \`\`\`
 
 **By role:**
-\`\`\`bash
+\`\`\`
 # Source mentions "their CTO"
-executeCommand("grep -r -i 'Role.*CTO' '{knowledge_folder}/People/'")
+workspace-grep({ pattern: "Role.*CTO", searchPath: "{knowledge_folder}/People" })
 # Filter results by organization context
 \`\`\`
 
@@ -844,7 +887,12 @@ If role is unknown but context suggests it, say so:
 
 One line summarizing this source's relevance to the entity:
 \`\`\`
-**{YYYY-MM-DD}** ({meeting|email}): {Summary with [[links]]}
+**{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[links]]}
+\`\`\`
+
+**For voice memos:** Include a link to the voice memo file using the Path field:
+\`\`\`
+**2025-01-15** (voice memo): Discussed [[Projects/Acme Integration]] timeline. See [[Voice Memos/2025-01-15/voice-memo-2025-01-15T10-30-00-000Z]]
 \`\`\`
 
 **Important:** Use canonical names with absolute paths from resolution map in all summaries:
@@ -968,8 +1016,8 @@ STATE CHANGES:
 Before writing, compare extracted content against existing notes.
 
 ## Check Activity Log
-\`\`\`bash
+\`\`\`
-executeCommand("grep '2025-01-15' '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-grep({ pattern: "2025-01-15", searchPath: "{knowledge_folder}/People/Sarah Chen.md" })
 \`\`\`
 
 If an entry for this date/source already exists, this may have been processed. Skip or verify different interaction.
@@ -999,28 +1047,28 @@ If new info contradicts existing:
 
 **IMPORTANT: Write sequentially, one file at a time.**
 - Generate content for exactly one note.
-- Issue exactly one \`write\` command.
+- Issue exactly one write/edit command.
 - Wait for the tool to return before generating the next note.
-- Do NOT batch multiple \`write\` commands in a single response.
+- Do NOT batch multiple write commands in a single response.
 
-**For new entities (meetings only):**
+**For NEW entities (use workspace-writeFile):**
-\`\`\`bash
+\`\`\`
-executeCommand("write '{knowledge_folder}/People/Jennifer.md' '{content}'")
+workspace-writeFile({
+  path: "{knowledge_folder}/People/Jennifer.md",
+  data: "# Jennifer\\n\\n## Summary\\n..."
+})
 \`\`\`
 
-**For existing entities:**
+**For EXISTING entities (use workspace-edit):**
-- Read current content first
+- Read current content first with workspace-readFile
-- Add activity entry at TOP of Activity section (reverse chronological)
+- Use workspace-edit to add activity entry at TOP (reverse chronological)
-- Update "Last seen" date
+- Update fields using targeted edits
-- Add new key facts (skip duplicates)
+\`\`\`
-- Add new open items
+workspace-edit({
-- Add new decisions
+  path: "{knowledge_folder}/People/Sarah Chen.md",
-- Add new relationships
+  oldString: "## Activity\\n",
-- Update summary ONLY if significant new understanding
+  newString: "## Activity\\n- **2026-02-03** (meeting): Met to discuss project timeline\\n"
-\`\`\`bash
+})
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
-# ... modify content ...
-executeCommand("write '{knowledge_folder}/People/Sarah Chen.md' '{full_updated_content}'")
 \`\`\`
 
 ## 9b: Emails — Update Existing Notes Only
@@ -1043,7 +1091,7 @@ For each state change identified in Step 7:
 ### Update Project Status
 \`\`\`bash
 # Read current project note
-executeCommand("cat '{knowledge_folder}/Projects/Acme Integration.md'")
+workspace-readFile({ path: "{knowledge_folder}/Projects/Acme Integration.md" })
 
 # Update the Status field
 # Change: **Status:** planning
@@ -1053,7 +1101,7 @@ executeCommand("cat '{knowledge_folder}/Projects/Acme Integration.md'")
 ### Mark Open Items Complete
 \`\`\`bash
 # Read current note
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-readFile({ path: "{knowledge_folder}/People/Sarah Chen.md" })
 
 # Find matching open item and update
 # Change: - [ ] Send API documentation — by Friday
@@ -1063,7 +1111,7 @@ executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
 ### Update Role
 \`\`\`bash
 # Read current person note
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-readFile({ path: "{knowledge_folder}/People/Sarah Chen.md" })
 
 # Update role field
 # Change: **Role:** Engineering Lead
@@ -1073,7 +1121,7 @@ executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
 ### Update Relationship
 \`\`\`bash
 # Read current org note
-executeCommand("cat '{knowledge_folder}/Organizations/Acme Corp.md'")
+workspace-readFile({ path: "{knowledge_folder}/Organizations/Acme Corp.md" })
 
 # Update relationship field
 # Change: **Relationship:** prospect
@@ -1138,8 +1186,8 @@ This ensures:
 ## Check Each New Link
 
 If you added \`[[People/Jennifer]]\` to \`Organizations/Acme Corp.md\`:
-\`\`\`bash
+\`\`\`
-executeCommand("grep 'Acme Corp' '{knowledge_folder}/People/Jennifer.md'")
+workspace-grep({ pattern: "Acme Corp", searchPath: "{knowledge_folder}/People/Jennifer.md" })
 \`\`\`
 
 If not found, update Jennifer.md to add the link.
@@ -1179,7 +1227,7 @@ If not found, update Jennifer.md to add the link.
 - [[Projects/{Project}]] — {role}
 
 ## Activity
-- **{YYYY-MM-DD}** ({meeting|email}): {Summary with [[Folder/Name]] links} {[State changes if any]}
+- **{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[Folder/Name]] links} {[State changes if any]}
 
 ## Key facts
 {Substantive facts only. Leave empty if none. Never include data gap commentary.}
@@ -1216,7 +1264,7 @@ If not found, update Jennifer.md to add the link.
 - [[Projects/{Project}]] — {relationship}
 
 ## Activity
-- **{YYYY-MM-DD}** ({meeting|email}): {Summary with [[Folder/Name]] links} {[State changes if any]}
+- **{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[Folder/Name]] links} {[State changes if any]}
 
 ## Key facts
 {Substantive facts only. Leave empty if none.}
@@ -1373,11 +1421,11 @@ Not mass email, not automated. Continue.
 - Variants: "Sarah Chen", "sarah@acme.com", "David Kim", "David", "Jennifer", "CTO", "Acme", "the pilot"
 
 ### Step 3: Search Existing Notes
-\`\`\`bash
+\`\`\`
-executeCommand("grep -r -i -l 'Sarah Chen' 'knowledge/'")
+workspace-grep({ pattern: "Sarah Chen", searchPath: "knowledge" })
 # Output: (none)
 
-executeCommand("grep -r -i -l 'acme' 'knowledge/'")
+workspace-grep({ pattern: "acme", searchPath: "knowledge" })
 # Output: (none)
 \`\`\`
 
@@ -1516,8 +1564,8 @@ VP Engineering, Acme Corp
 ### Step 1: Filter
 
 Check for existing relationship:
-\`\`\`bash
+\`\`\`
-executeCommand("grep -r -i -l 'sarah@acme.com' 'knowledge/'")
+workspace-grep({ pattern: "sarah@acme.com", searchPath: "knowledge" })
 # Output: notes/People/Sarah Chen.md
 \`\`\`
 
@@ -1652,11 +1700,11 @@ John Smith
 ### Step 1: Filter
 
 Check for existing relationship:
-\`\`\`bash
+\`\`\`
-executeCommand("grep -r -i -l 'randomvendor' 'knowledge/'")
+workspace-grep({ pattern: "randomvendor", searchPath: "knowledge" })
 # Output: (none)
 
-executeCommand("grep -r -i -l 'John Smith' 'knowledge/'")
+workspace-grep({ pattern: "John Smith", searchPath: "knowledge" })
 # Output: (none)
 \`\`\`
 
@@ -1699,8 +1747,8 @@ David
 ### Step 1: Filter
 
 Check for sender:
-\`\`\`bash
+\`\`\`
-executeCommand("grep -r -i -l 'david@friendly.vc' 'knowledge/'")
+workspace-grep({ pattern: "david@friendly.vc", searchPath: "knowledge" })
 # Output: notes/People/David Park.md
 \`\`\`
 
@@ -1803,10 +1851,16 @@ Business banking provider. Account setup completed January 2025.
 | Source Type | Creates Notes? | Updates Notes? | Detects State Changes? |
 |-------------|---------------|----------------|------------------------|
 | Meeting | Yes | Yes | Yes |
+| Voice memo | Yes | Yes | Yes |
 | Email (known contact) | No | Yes | Yes |
 | Email (unknown contact) | No | No (SKIP) | No |
 | Email (warm intro) | Yes (exception) | Yes | Yes |
 
+**Voice memo activity format:** Always include a link to the source voice memo:
+\`\`\`
+**2025-01-15** (voice memo): Discussed project timeline with [[People/Sarah Chen]]. See [[Voice Memos/2025-01-15/voice-memo-...]]
+\`\`\`
+
 ---
 
 # State Change Reference

apps/x/packages/core/src/knowledge/note_creation_low.ts

@@ -7,19 +7,25 @@ tools:
   workspace-readFile:
     type: builtin
     name: workspace-readFile
+  workspace-edit:
+    type: builtin
+    name: workspace-edit
   workspace-readdir:
     type: builtin
     name: workspace-readdir
   workspace-mkdir:
     type: builtin
     name: workspace-mkdir
-  executeCommand:
+  workspace-grep:
     type: builtin
-    name: executeCommand
+    name: workspace-grep
+  workspace-glob:
+    type: builtin
+    name: workspace-glob
 ---
 # Task
 
-You are a memory agent. Given a single source file (email or meeting transcript), you will:
+You are a memory agent. Given a single source file (email, meeting transcript, or voice memo), you will:
 
 1. **Determine source type (meeting or email)**
 2. **Evaluate if the source is worth processing**
@@ -31,7 +37,7 @@ You are a memory agent. Given a single source file (email or meeting transcript)
 8. Create new notes or update existing notes
 9. **Apply state changes to existing notes**
 
-The core rule: **Capture broadly. Both meetings and emails create notes for most external contacts.**
+The core rule: **Capture broadly. Meetings, voice memos, and emails create notes for most external contacts.**
 
 You have full read access to the existing knowledge directory. Use this extensively to:
 - Find existing notes for people, organizations, projects mentioned
@@ -70,20 +76,51 @@ When you need to:
 
 # Tools Available
 
-You have access to \`executeCommand\` to run shell commands:
+You have access to these tools:
+
+**For reading files:**
 \`\`\`
-executeCommand("ls {path}")                     # List directory contents
+workspace-readFile({ path: "knowledge/People/Sarah Chen.md" })
-executeCommand("cat {path}")                    # Read file contents
-executeCommand("head -50 {path}")               # Read first 50 lines
-executeCommand("write {path} {content}")        # Create or overwrite file
 \`\`\`
 
-**Important:** Use shell escaping for paths with spaces:
+**For creating NEW files:**
 \`\`\`
-executeCommand("cat 'knowledge_folder/People/Sarah Chen.md'")
+workspace-writeFile({ path: "knowledge/People/Sarah Chen.md", data: "# Sarah Chen\\n\\n..." })
 \`\`\`
 
-**NOTE:** Do NOT use grep to search for entities. Use the provided knowledge_index instead.
+**For editing EXISTING files (preferred for updates):**
+\`\`\`
+workspace-edit({
+  path: "knowledge/People/Sarah Chen.md",
+  oldString: "## Activity\\n",
+  newString: "## Activity\\n- **2026-02-03** (meeting): New activity entry\\n"
+})
+\`\`\`
+
+**For listing directories:**
+\`\`\`
+workspace-readdir({ path: "knowledge/People" })
+\`\`\`
+
+**For creating directories:**
+\`\`\`
+workspace-mkdir({ path: "knowledge/Projects", recursive: true })
+\`\`\`
+
+**For searching files:**
+\`\`\`
+workspace-grep({ pattern: "Acme Corp", searchPath: "knowledge", fileGlob: "*.md" })
+\`\`\`
+
+**For finding files by pattern:**
+\`\`\`
+workspace-glob({ pattern: "**/*.md", cwd: "knowledge/People" })
+\`\`\`
+
+**IMPORTANT:**
+- Use \`workspace-edit\` for updating existing notes (adding activity, updating fields)
+- Use \`workspace-writeFile\` only for creating new notes
+- Prefer the knowledge_index for entity resolution (it's faster than grep)
 
 # Output
 
@@ -120,7 +157,7 @@ This mode prioritizes comprehensive capture over selectivity. The goal is to nev
 
 Read the source file and determine if it's a meeting or email.
 \`\`\`
-executeCommand("cat '{source_file}'")
+workspace-readFile({ path: "{source_file}" })
 \`\`\`
 
 **Meeting indicators:**
@@ -133,9 +170,15 @@ executeCommand("cat '{source_file}'")
 - Has \`Subject:\` field
 - Email signature
 
+**Voice memo indicators:**
+- Has \`**Type:** voice memo\` field
+- Has \`**Path:**\` field with path like \`Voice Memos/YYYY-MM-DD/...\`
+- Has \`## Transcript\` section
+
 **Set processing mode:**
 - \`source_type = "meeting"\` → Create notes for all external attendees
 - \`source_type = "email"\` → Create notes for sender if identifiable human
+- \`source_type = "voice_memo"\` → Create notes for all mentioned entities (treat like meetings)
 
 ---
 
@@ -204,7 +247,7 @@ If processing, continue to Step 2.
 
 # Step 2: Read and Parse Source File
 \`\`\`
-executeCommand("cat '{source_file}'")
+workspace-readFile({ path: "{source_file}" })
 \`\`\`
 
 Extract metadata:
@@ -291,8 +334,8 @@ From index, find matches for:
 ## 3d: Read Full Notes When Needed
 
 Only read the full note content when you need details not in the index (e.g., activity logs, open items):
-\`\`\`bash
+\`\`\`
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-readFile({ path: "{knowledge_folder}/People/Sarah Chen.md" })
 \`\`\`
 
 **Why read these notes:**
@@ -509,7 +552,12 @@ Write 2-3 sentences covering their role/function, context of the relationship, a
 
 One line summarizing this source's relevance to the entity:
 \`\`\`
-**{YYYY-MM-DD}** ({meeting|email}): {Summary with [[links]]}
+**{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[links]]}
+\`\`\`
+
+**For voice memos:** Include a link to the voice memo file using the Path field:
+\`\`\`
+**2025-01-15** (voice memo): Discussed [[Projects/Acme Integration]] timeline. See [[Voice Memos/2025-01-15/voice-memo-2025-01-15T10-30-00-000Z]]
 \`\`\`
 
 ---
@@ -555,21 +603,29 @@ Before writing:
 
 **IMPORTANT: Write sequentially, one file at a time.**
 - Generate content for exactly one note.
-- Issue exactly one \`write\` command.
+- Issue exactly one write/edit command.
 - Wait for the tool to return before generating the next note.
-- Do NOT batch multiple \`write\` commands in a single response.
+- Do NOT batch multiple write commands in a single response.
 
-**For new entities:**
+**For NEW entities (use workspace-writeFile):**
-\`\`\`bash
+\`\`\`
-executeCommand("write '{knowledge_folder}/People/Jennifer.md' '{content}'")
+workspace-writeFile({
+  path: "{knowledge_folder}/People/Jennifer.md",
+  data: "# Jennifer\\n\\n## Summary\\n..."
+})
 \`\`\`
 
-**For existing entities:**
+**For EXISTING entities (use workspace-edit):**
-- Read current content first
+- Read current content first with workspace-readFile
-- Add activity entry at TOP (reverse chronological)
+- Use workspace-edit to add activity entry at TOP (reverse chronological)
-- Update "Last seen" date
+- Update fields using targeted edits
-- Add new key facts (skip duplicates)
+\`\`\`
-- Add new open items
+workspace-edit({
+  path: "{knowledge_folder}/People/Sarah Chen.md",
+  oldString: "## Activity\\n",
+  newString: "## Activity\\n- **2026-02-03** (meeting): Met to discuss project timeline\\n"
+})
+\`\`\`
 
 ## 9b: Apply State Changes
 
@@ -638,7 +694,7 @@ After writing, verify links go both ways.
 - [[Projects/{Project}]] — {role}
 
 ## Activity
-- **{YYYY-MM-DD}** ({meeting|email}): {Summary with [[Folder/Name]] links}
+- **{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[Folder/Name]] links}
 
 ## Key facts
 {Substantive facts only. Leave empty if none.}
@@ -673,7 +729,7 @@ After writing, verify links go both ways.
 - [[Projects/{Project}]] — {relationship}
 
 ## Activity
-- **{YYYY-MM-DD}** ({meeting|email}): {Summary}
+- **{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary}
 
 ## Key facts
 {Substantive facts only. Leave empty if none.}
@@ -705,7 +761,7 @@ After writing, verify links go both ways.
 - [[Topics/{Topic}]] — {relationship}
 
 ## Timeline
-**{YYYY-MM-DD}** ({meeting|email})
+**{YYYY-MM-DD}** ({meeting|email|voice memo})
 {What happened.}
 
 ## Decisions
@@ -756,9 +812,15 @@ After writing, verify links go both ways.
 | Source Type | Creates Notes? | Updates Notes? | Detects State Changes? |
 |-------------|---------------|----------------|------------------------|
 | Meeting | Yes — ALL external attendees | Yes | Yes |
+| Voice memo | Yes — all mentioned entities | Yes | Yes |
 | Email (any human sender) | Yes | Yes | Yes |
 | Email (automated/newsletter) | No (SKIP) | No | No |
 
+**Voice memo activity format:** Always include a link to the source voice memo:
+\`\`\`
+**2025-01-15** (voice memo): Discussed project timeline with [[People/Sarah Chen]]. See [[Voice Memos/2025-01-15/voice-memo-...]]
+\`\`\`
+
 **Philosophy:** Capture broadly, filter later if needed.
 
 ---

apps/x/packages/core/src/knowledge/note_creation_medium.ts

@@ -7,19 +7,25 @@ tools:
   workspace-readFile:
     type: builtin
     name: workspace-readFile
+  workspace-edit:
+    type: builtin
+    name: workspace-edit
   workspace-readdir:
     type: builtin
     name: workspace-readdir
   workspace-mkdir:
     type: builtin
     name: workspace-mkdir
-  executeCommand:
+  workspace-grep:
     type: builtin
-    name: executeCommand
+    name: workspace-grep
+  workspace-glob:
+    type: builtin
+    name: workspace-glob
 ---
 # Task
 
-You are a memory agent. Given a single source file (email or meeting transcript), you will:
+You are a memory agent. Given a single source file (email, meeting transcript, or voice memo), you will:
 
 1. **Determine source type (meeting or email)**
 2. **Evaluate if the source is worth processing**
@@ -70,20 +76,51 @@ When you need to:
 
 # Tools Available
 
-You have access to \`executeCommand\` to run shell commands:
+You have access to these tools:
+
+**For reading files:**
 \`\`\`
-executeCommand("ls {path}")                     # List directory contents
+workspace-readFile({ path: "knowledge/People/Sarah Chen.md" })
-executeCommand("cat {path}")                    # Read file contents
-executeCommand("head -50 {path}")               # Read first 50 lines
-executeCommand("write {path} {content}")        # Create or overwrite file
 \`\`\`
 
-**Important:** Use shell escaping for paths with spaces:
+**For creating NEW files:**
 \`\`\`
-executeCommand("cat 'knowledge_folder/People/Sarah Chen.md'")
+workspace-writeFile({ path: "knowledge/People/Sarah Chen.md", data: "# Sarah Chen\\n\\n..." })
 \`\`\`
 
-**NOTE:** Do NOT use grep to search for entities. Use the provided knowledge_index instead.
+**For editing EXISTING files (preferred for updates):**
+\`\`\`
+workspace-edit({
+  path: "knowledge/People/Sarah Chen.md",
+  oldString: "## Activity\\n",
+  newString: "## Activity\\n- **2026-02-03** (meeting): New activity entry\\n"
+})
+\`\`\`
+
+**For listing directories:**
+\`\`\`
+workspace-readdir({ path: "knowledge/People" })
+\`\`\`
+
+**For creating directories:**
+\`\`\`
+workspace-mkdir({ path: "knowledge/Projects", recursive: true })
+\`\`\`
+
+**For searching files:**
+\`\`\`
+workspace-grep({ pattern: "Acme Corp", searchPath: "knowledge", fileGlob: "*.md" })
+\`\`\`
+
+**For finding files by pattern:**
+\`\`\`
+workspace-glob({ pattern: "**/*.md", cwd: "knowledge/People" })
+\`\`\`
+
+**IMPORTANT:**
+- Use \`workspace-edit\` for updating existing notes (adding activity, updating fields)
+- Use \`workspace-writeFile\` only for creating new notes
+- Prefer the knowledge_index for entity resolution (it's faster than grep)
 
 # Output
 
@@ -119,7 +156,7 @@ Either:
 
 Read the source file and determine if it's a meeting or email.
 \`\`\`
-executeCommand("cat '{source_file}'")
+workspace-readFile({ path: "{source_file}" })
 \`\`\`
 
 **Meeting indicators:**
@@ -132,9 +169,15 @@ executeCommand("cat '{source_file}'")
 - Has \`Subject:\` field
 - Email signature
 
+**Voice memo indicators:**
+- Has \`**Type:** voice memo\` field
+- Has \`**Path:**\` field with path like \`Voice Memos/YYYY-MM-DD/...\`
+- Has \`## Transcript\` section
+
 **Set processing mode:**
 - \`source_type = "meeting"\` → Can create new notes
 - \`source_type = "email"\` → Can create notes if personalized and relevant
+- \`source_type = "voice_memo"\` → Can create new notes (treat like meetings)
 
 ---
 
@@ -344,7 +387,7 @@ If processing, continue to Step 2.
 
 # Step 2: Read and Parse Source File
 \`\`\`
-executeCommand("cat '{source_file}'")
+workspace-readFile({ path: "{source_file}" })
 \`\`\`
 
 Extract metadata:
@@ -441,7 +484,7 @@ From index, find matches for:
 
 Only read the full note content when you need details not in the index (e.g., activity logs, open items):
 \`\`\`bash
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-readFile({ path: "{knowledge_folder}/People/Sarah Chen.md" })
 \`\`\`
 
 **Why read these notes:**
@@ -525,27 +568,27 @@ Resolution Map:
 When multiple candidates match a variant, disambiguate:
 
 **By organization (strongest signal):**
-\`\`\`bash
+\`\`\`
 # "David" could be David Kim or David Chen
-executeCommand("grep -i 'Acme' '{knowledge_folder}/People/David Kim.md'")
+workspace-grep({ pattern: "Acme", searchPath: "{knowledge_folder}/People/David Kim.md" })
 # Output: **Organization:** [[Acme Corp]]
 
-executeCommand("grep -i 'Acme' '{knowledge_folder}/People/David Chen.md'")
+workspace-grep({ pattern: "Acme", searchPath: "{knowledge_folder}/People/David Chen.md" })
 # Output: **Organization:** [[Other Corp]]
 
 # Source is from Acme context → "David" = "David Kim"
 \`\`\`
 
 **By email (definitive):**
-\`\`\`bash
+\`\`\`
-executeCommand("grep -i 'david@acme.com' '{knowledge_folder}/People/David Kim.md'")
+workspace-grep({ pattern: "david@acme.com", searchPath: "{knowledge_folder}/People/David Kim.md" })
 # Exact email match is definitive
 \`\`\`
 
 **By role:**
-\`\`\`bash
+\`\`\`
 # Source mentions "their CTO"
-executeCommand("grep -r -i 'Role.*CTO' '{knowledge_folder}/People/'")
+workspace-grep({ pattern: "Role.*CTO", searchPath: "{knowledge_folder}/People" })
 # Filter results by organization context
 \`\`\`
 
@@ -784,7 +827,12 @@ The summary should answer: **"Who is this person and why do I know them?"**
 
 One line summarizing this source's relevance to the entity:
 \`\`\`
-**{YYYY-MM-DD}** ({meeting|email}): {Summary with [[links]]}
+**{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[links]]}
+\`\`\`
+
+**For voice memos:** Include a link to the voice memo file using the Path field:
+\`\`\`
+**2025-01-15** (voice memo): Discussed [[Projects/Acme Integration]] timeline. See [[Voice Memos/2025-01-15/voice-memo-2025-01-15T10-30-00-000Z]]
 \`\`\`
 
 **Important:** Use canonical names with absolute paths from resolution map in all summaries:
@@ -877,8 +925,8 @@ STATE CHANGES:
 Before writing, compare extracted content against existing notes.
 
 ## Check Activity Log
-\`\`\`bash
+\`\`\`
-executeCommand("grep '2025-01-15' '{knowledge_folder}/People/Sarah Chen.md'")
+workspace-grep({ pattern: "2025-01-15", searchPath: "{knowledge_folder}/People/Sarah Chen.md" })
 \`\`\`
 
 If an entry for this date/source already exists, this may have been processed. Skip or verify different interaction.
@@ -908,28 +956,28 @@ If new info contradicts existing:
 
 **IMPORTANT: Write sequentially, one file at a time.**
 - Generate content for exactly one note.
-- Issue exactly one \`write\` command.
+- Issue exactly one write/edit command.
 - Wait for the tool to return before generating the next note.
-- Do NOT batch multiple \`write\` commands in a single response.
+- Do NOT batch multiple write commands in a single response.
 
-**For new entities (meetings and qualifying emails):**
+**For NEW entities (use workspace-writeFile):**
-\`\`\`bash
+\`\`\`
-executeCommand("write '{knowledge_folder}/People/Jennifer.md' '{content}'")
+workspace-writeFile({
+  path: "{knowledge_folder}/People/Jennifer.md",
+  data: "# Jennifer\\n\\n## Summary\\n..."
+})
 \`\`\`
 
-**For existing entities:**
+**For EXISTING entities (use workspace-edit):**
-- Read current content first
+- Read current content first with workspace-readFile
-- Add activity entry at TOP of Activity section (reverse chronological)
+- Use workspace-edit to add activity entry at TOP (reverse chronological)
-- Update "Last seen" date
+- Update fields using targeted edits
-- Add new key facts (skip duplicates)
+\`\`\`
-- Add new open items
+workspace-edit({
-- Add new decisions
+  path: "{knowledge_folder}/People/Sarah Chen.md",
-- Add new relationships
+  oldString: "## Activity\\n",
-- Update summary ONLY if significant new understanding
+  newString: "## Activity\\n- **2026-02-03** (meeting): Met to discuss project timeline\\n"
-\`\`\`bash
+})
-executeCommand("cat '{knowledge_folder}/People/Sarah Chen.md'")
-# ... modify content ...
-executeCommand("write '{knowledge_folder}/People/Sarah Chen.md' '{full_updated_content}'")
 \`\`\`
 
 ## 9b: Apply State Changes
@@ -1000,7 +1048,7 @@ After writing, verify links go both ways.
 - [[Projects/{Project}]] — {role}
 
 ## Activity
-- **{YYYY-MM-DD}** ({meeting|email}): {Summary with [[Folder/Name]] links}
+- **{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[Folder/Name]] links}
 
 ## Key facts
 {Substantive facts only. Leave empty if none.}
@@ -1035,7 +1083,7 @@ After writing, verify links go both ways.
 - [[Projects/{Project}]] — {relationship}
 
 ## Activity
-- **{YYYY-MM-DD}** ({meeting|email}): {Summary with [[Folder/Name]] links}
+- **{YYYY-MM-DD}** ({meeting|email|voice memo}): {Summary with [[Folder/Name]] links}
 
 ## Key facts
 {Substantive facts only. Leave empty if none.}
@@ -1119,11 +1167,17 @@ After writing, verify links go both ways.
 | Source Type | Creates Notes? | Updates Notes? | Detects State Changes? |
 |-------------|---------------|----------------|------------------------|
 | Meeting | Yes | Yes | Yes |
+| Voice memo | Yes | Yes | Yes |
 | Email (personalized, business-relevant) | Yes | Yes | Yes |
 | Email (mass/automated/consumer) | No (SKIP) | No | No |
 | Email (cold outreach with personalization) | Yes | Yes | Yes |
 | Email (generic cold outreach) | No | No | No |
 
+**Voice memo activity format:** Always include a link to the source voice memo:
+\`\`\`
+**2025-01-15** (voice memo): Discussed project timeline with [[People/Sarah Chen]]. See [[Voice Memos/2025-01-15/voice-memo-...]]
+\`\`\`
+
 ---
 
 # Error Handling