kimaki 0.4.35 → 0.4.37

package/src/message-formatting.ts

@@ -77,8 +77,8 @@ export function isTextMimeType(contentType: string | null): boolean {
 }
 
 export async function getTextAttachments(message: Message): Promise<string> {
-  const textAttachments = Array.from(message.attachments.values()).filter(
-    (attachment) => isTextMimeType(attachment.contentType),
+  const textAttachments = Array.from(message.attachments.values()).filter((attachment) =>
+    isTextMimeType(attachment.contentType),
   )
 
   if (textAttachments.length === 0) {
@@ -105,14 +105,10 @@ export async function getTextAttachments(message: Message): Promise<string> {
 }
 
 export async function getFileAttachments(message: Message): Promise<FilePartInput[]> {
-  const fileAttachments = Array.from(message.attachments.values()).filter(
-    (attachment) => {
-      const contentType = attachment.contentType || ''
-      return (
-        contentType.startsWith('image/') || contentType === 'application/pdf'
-      )
-    },
-  )
+  const fileAttachments = Array.from(message.attachments.values()).filter((attachment) => {
+    const contentType = attachment.contentType || ''
+    return contentType.startsWith('image/') || contentType === 'application/pdf'
+  })
 
   if (fileAttachments.length === 0) {
     return []
@@ -164,7 +160,9 @@ export function getToolSummaryText(part: Part): string {
     const added = newString.split('\n').length
     const removed = oldString.split('\n').length
     const fileName = filePath.split('/').pop() || ''
-    return fileName ? `*${escapeInlineMarkdown(fileName)}* (+${added}-${removed})` : `(+${added}-${removed})`
+    return fileName
+      ? `*${escapeInlineMarkdown(fileName)}* (+${added}-${removed})`
+      : `(+${added}-${removed})`
   }
 
   if (part.tool === 'write') {
@@ -172,7 +170,9 @@ export function getToolSummaryText(part: Part): string {
     const content = (part.state.input?.content as string) || ''
     const lines = content.split('\n').length
     const fileName = filePath.split('/').pop() || ''
-    return fileName ? `*${escapeInlineMarkdown(fileName)}* (${lines} line${lines === 1 ? '' : 's'})` : `(${lines} line${lines === 1 ? '' : 's'})`
+    return fileName
+      ? `*${escapeInlineMarkdown(fileName)}* (${lines} line${lines === 1 ? '' : 's'})`
+      : `(${lines} line${lines === 1 ? '' : 's'})`
   }
 
   if (part.tool === 'webfetch') {
@@ -259,8 +259,7 @@ export function formatPart(part: Part): string {
     const trimmed = part.text.trimStart()
     const firstChar = trimmed[0] || ''
     const markdownStarters = ['#', '*', '_', '-', '>', '`', '[', '|']
-    const startsWithMarkdown =
-      markdownStarters.includes(firstChar) || /^\d+\./.test(trimmed)
+    const startsWithMarkdown = markdownStarters.includes(firstChar) || /^\d+\./.test(trimmed)
     if (startsWithMarkdown) {
       return `\n${part.text}`
     }

package/src/openai-realtime.ts

@@ -129,13 +129,7 @@ function createWavHeader(dataLength: number, options: WavConversionOptions) {
   return buffer
 }
 
-function defaultAudioChunkHandler({
-  data,
-  mimeType,
-}: {
-  data: Buffer
-  mimeType: string
-}) {
+function defaultAudioChunkHandler({ data, mimeType }: { data: Buffer; mimeType: string }) {
   audioParts.push(data)
   const fileName = 'audio.wav'
   const buffer = convertToWav(audioParts, mimeType)
@@ -247,36 +241,23 @@ export async function startGenAiSession({
   }
 
   // Set up event handlers
-  client.on(
-    'conversation.item.created',
-    ({ item }: { item: ConversationItem }) => {
-      if (
-        'role' in item &&
-        item.role === 'assistant' &&
-        item.type === 'message'
-      ) {
-        // Check if this is the first audio content
-        const hasAudio =
-          'content' in item &&
-          Array.isArray(item.content) &&
-          item.content.some((c) => 'type' in c && c.type === 'audio')
-        if (hasAudio && !isAssistantSpeaking && onAssistantStartSpeaking) {
-          isAssistantSpeaking = true
-          onAssistantStartSpeaking()
-        }
+  client.on('conversation.item.created', ({ item }: { item: ConversationItem }) => {
+    if ('role' in item && item.role === 'assistant' && item.type === 'message') {
+      // Check if this is the first audio content
+      const hasAudio =
+        'content' in item &&
+        Array.isArray(item.content) &&
+        item.content.some((c) => 'type' in c && c.type === 'audio')
+      if (hasAudio && !isAssistantSpeaking && onAssistantStartSpeaking) {
+        isAssistantSpeaking = true
+        onAssistantStartSpeaking()
       }
-    },
-  )
+    }
+  })
 
   client.on(
     'conversation.updated',
-    ({
-      item,
-      delta,
-    }: {
-      item: ConversationItem
-      delta: ConversationEventDelta | null
-    }) => {
+    ({ item, delta }: { item: ConversationItem; delta: ConversationEventDelta | null }) => {
       // Handle audio chunks
       if (delta?.audio && 'role' in item && item.role === 'assistant') {
         if (!isAssistantSpeaking && onAssistantStartSpeaking) {
@@ -313,20 +294,17 @@ export async function startGenAiSession({
     },
   )
 
-  client.on(
-    'conversation.item.completed',
-    ({ item }: { item: ConversationItem }) => {
-      if (
-        'role' in item &&
-        item.role === 'assistant' &&
-        isAssistantSpeaking &&
-        onAssistantStopSpeaking
-      ) {
-        isAssistantSpeaking = false
-        onAssistantStopSpeaking()
-      }
-    },
-  )
+  client.on('conversation.item.completed', ({ item }: { item: ConversationItem }) => {
+    if (
+      'role' in item &&
+      item.role === 'assistant' &&
+      isAssistantSpeaking &&
+      onAssistantStopSpeaking
+    ) {
+      isAssistantSpeaking = false
+      onAssistantStopSpeaking()
+    }
+  })
 
   client.on('conversation.interrupted', () => {
     openaiLogger.log('Assistant was interrupted')

package/src/opencode.ts

@@ -5,11 +5,7 @@
 import { spawn, type ChildProcess } from 'node:child_process'
 import fs from 'node:fs'
 import net from 'node:net'
-import {
-  createOpencodeClient,
-  type OpencodeClient,
-  type Config,
-} from '@opencode-ai/sdk'
+import { createOpencodeClient, type OpencodeClient, type Config } from '@opencode-ai/sdk'
 import {
   createOpencodeClient as createOpencodeClientV2,
   type OpencodeClient as OpencodeClientV2,
@@ -84,9 +80,7 @@ async function waitForServer(port: number, maxAttempts = 30): Promise<boolean> {
     }
     await new Promise((resolve) => setTimeout(resolve, 1000))
   }
-  throw new Error(
-    `Server did not start on port ${port} after ${maxAttempts} seconds`,
-  )
+  throw new Error(`Server did not start on port ${port} after ${maxAttempts} seconds`)
 }
 
 export async function initializeOpencodeForDirectory(directory: string) {
@@ -117,33 +111,31 @@ export async function initializeOpencodeForDirectory(directory: string) {
 
   const opencodeCommand = process.env.OPENCODE_PATH || 'opencode'
 
-  const serverProcess = spawn(
-    opencodeCommand,
-    ['serve', '--port', port.toString()],
-    {
-      stdio: 'pipe',
-      detached: false,
-      cwd: directory,
-      env: {
-        ...process.env,
-        OPENCODE_CONFIG_CONTENT: JSON.stringify({
-          $schema: 'https://opencode.ai/config.json',
-          lsp: false,
-          formatter: false,
-          permission: {
-            edit: 'allow',
-            bash: 'allow',
-            webfetch: 'allow',
-          },
-        } satisfies Config),
-        OPENCODE_PORT: port.toString(),
-      },
+  const serverProcess = spawn(opencodeCommand, ['serve', '--port', port.toString()], {
+    stdio: 'pipe',
+    detached: false,
+    cwd: directory,
+    env: {
+      ...process.env,
+      OPENCODE_CONFIG_CONTENT: JSON.stringify({
+        $schema: 'https://opencode.ai/config.json',
+        lsp: false,
+        formatter: false,
+        permission: {
+          edit: 'allow',
+          bash: 'allow',
+          webfetch: 'allow',
+        },
+      } satisfies Config),
+      OPENCODE_PORT: port.toString(),
     },
-  )
+  })
 
   // Buffer logs until we know if server started successfully
   const logBuffer: string[] = []
-  logBuffer.push(`Spawned opencode serve --port ${port} in ${directory} (pid: ${serverProcess.pid})`)
+  logBuffer.push(
+    `Spawned opencode serve --port ${port} in ${directory} (pid: ${serverProcess.pid})`,
+  )
 
   serverProcess.stdout?.on('data', (data) => {
     logBuffer.push(`[stdout] ${data.toString().trim()}`)
@@ -171,9 +163,7 @@ export async function initializeOpencodeForDirectory(directory: string) {
           opencodeLogger.error(`Failed to restart opencode server:`, e)
         })
       } else {
-        opencodeLogger.error(
-          `Server for ${directory} crashed too many times (5), not restarting`,
-        )
+        opencodeLogger.error(`Server for ${directory} crashed too many times (5), not restarting`)
       }
     } else {
       serverRetryCount.delete(directory)

package/src/session-handler.ts

@@ -6,14 +6,29 @@ import type { Part, PermissionRequest } from '@opencode-ai/sdk/v2'
 import type { FilePartInput } from '@opencode-ai/sdk'
 import type { Message, ThreadChannel } from 'discord.js'
 import prettyMilliseconds from 'pretty-ms'
-import { getDatabase, getSessionModel, getChannelModel, getSessionAgent, getChannelAgent, setSessionAgent } from './database.js'
-import { initializeOpencodeForDirectory, getOpencodeServers, getOpencodeClientV2 } from './opencode.js'
+import {
+  getDatabase,
+  getSessionModel,
+  getChannelModel,
+  getSessionAgent,
+  getChannelAgent,
+  setSessionAgent,
+} from './database.js'
+import {
+  initializeOpencodeForDirectory,
+  getOpencodeServers,
+  getOpencodeClientV2,
+} from './opencode.js'
 import { sendThreadMessage, NOTIFY_MESSAGE_FLAGS, SILENT_MESSAGE_FLAGS } from './discord-utils.js'
 import { formatPart } from './message-formatting.js'
 import { getOpencodeSystemMessage } from './system-message.js'
 import { createLogger } from './logger.js'
 import { isAbortError } from './utils.js'
-import { showAskUserQuestionDropdowns, cancelPendingQuestion, pendingQuestionContexts } from './commands/ask-question.js'
+import {
+  showAskUserQuestionDropdowns,
+  cancelPendingQuestion,
+  pendingQuestionContexts,
+} from './commands/ask-question.js'
 import { showPermissionDropdown, cleanupPermissionContext } from './commands/permissions.js'
 
 const sessionLogger = createLogger('SESSION')
@@ -96,7 +111,9 @@ export async function abortAndRetrySession({
   }
 
   // Small delay to let the abort propagate
-  await new Promise((resolve) => { setTimeout(resolve, 300) })
+  await new Promise((resolve) => {
+    setTimeout(resolve, 300)
+  })
 
   // Fetch last user message from API
   sessionLogger.log(`[ABORT+RETRY] Fetching last user message for session ${sessionId}`)
@@ -110,7 +127,9 @@ export async function abortAndRetrySession({
   }
 
   // Extract text and images from parts
-  const textPart = lastUserMessage.parts.find((p) => p.type === 'text') as { type: 'text'; text: string } | undefined
+  const textPart = lastUserMessage.parts.find((p) => p.type === 'text') as
+    | { type: 'text'; text: string }
+    | undefined
   const prompt = textPart?.text || ''
   const images = lastUserMessage.parts.filter((p) => p.type === 'file') as FilePartInput[]
 
@@ -183,17 +202,13 @@ export async function handleOpencodeSession({
       session = sessionResponse.data
       sessionLogger.log(`Successfully reused session ${sessionId}`)
     } catch (error) {
-      voiceLogger.log(
-        `[SESSION] Session ${sessionId} not found, will create new one`,
-      )
+      voiceLogger.log(`[SESSION] Session ${sessionId} not found, will create new one`)
     }
   }
 
   if (!session) {
     const sessionTitle = prompt.length > 80 ? prompt.slice(0, 77) + '...' : prompt.slice(0, 80)
-    voiceLogger.log(
-      `[SESSION] Creating new session with title: "${sessionTitle}"`,
-    )
+    voiceLogger.log(`[SESSION] Creating new session with title: "${sessionTitle}"`)
     const sessionResponse = await getClient().session.create({
       body: { title: sessionTitle },
     })
@@ -206,9 +221,7 @@ export async function handleOpencodeSession({
   }
 
   getDatabase()
-    .prepare(
-      'INSERT OR REPLACE INTO thread_sessions (thread_id, session_id) VALUES (?, ?)',
-    )
+    .prepare('INSERT OR REPLACE INTO thread_sessions (thread_id, session_id) VALUES (?, ?)')
     .run(thread.id, session.id)
   sessionLogger.log(`Stored session ${session.id} for thread ${thread.id}`)
 
@@ -220,16 +233,16 @@ export async function handleOpencodeSession({
 
   const existingController = abortControllers.get(session.id)
   if (existingController) {
-    voiceLogger.log(
-      `[ABORT] Cancelling existing request for session: ${session.id}`,
-    )
+    voiceLogger.log(`[ABORT] Cancelling existing request for session: ${session.id}`)
     existingController.abort(new Error('New request started'))
   }
 
   const pendingPerm = pendingPermissions.get(thread.id)
   if (pendingPerm) {
     try {
-      sessionLogger.log(`[PERMISSION] Auto-rejecting pending permission ${pendingPerm.permission.id} due to new message`)
+      sessionLogger.log(
+        `[PERMISSION] Auto-rejecting pending permission ${pendingPerm.permission.id} due to new message`,
+      )
       const clientV2 = getOpencodeClientV2(directory)
       if (clientV2) {
         await clientV2.permission.reply({
@@ -240,7 +253,10 @@ export async function handleOpencodeSession({
       // Clean up both the pending permission and its dropdown context
       cleanupPermissionContext(pendingPerm.contextHash)
       pendingPermissions.delete(thread.id)
-      await sendThreadMessage(thread, `⚠️ Previous permission request auto-rejected due to new message`)
+      await sendThreadMessage(
+        thread,
+        `⚠️ Previous permission request auto-rejected due to new message`,
+      )
     } catch (e) {
       sessionLogger.log(`[PERMISSION] Failed to auto-reject permission:`, e)
       cleanupPermissionContext(pendingPerm.contextHash)
@@ -258,7 +274,9 @@ export async function handleOpencodeSession({
   abortControllers.set(session.id, abortController)
 
   if (existingController) {
-    await new Promise((resolve) => { setTimeout(resolve, 200) })
+    await new Promise((resolve) => {
+      setTimeout(resolve, 200)
+    })
     if (abortController.signal.aborted) {
       sessionLogger.log(`[DEBOUNCE] Request was superseded during wait, exiting`)
       return
@@ -277,7 +295,7 @@ export async function handleOpencodeSession({
   }
   const eventsResult = await clientV2.event.subscribe(
     { directory },
-    { signal: abortController.signal }
+    { signal: abortController.signal },
   )
 
   if (abortController.signal.aborted) {
@@ -289,10 +307,11 @@ export async function handleOpencodeSession({
   sessionLogger.log(`Subscribed to OpenCode events`)
 
   const sentPartIds = new Set<string>(
-    (getDatabase()
-      .prepare('SELECT part_id FROM part_messages WHERE thread_id = ?')
-      .all(thread.id) as { part_id: string }[])
-      .map((row) => row.part_id)
+    (
+      getDatabase()
+        .prepare('SELECT part_id FROM part_messages WHERE thread_id = ?')
+        .all(thread.id) as { part_id: string }[]
+    ).map((row) => row.part_id),
   )
 
   let currentParts: Part[] = []
@@ -385,7 +404,12 @@ export async function handleOpencodeSession({
           }
 
           if (msg.role === 'assistant') {
-            const newTokensTotal = msg.tokens.input + msg.tokens.output + msg.tokens.reasoning + msg.tokens.cache.read + msg.tokens.cache.write
+            const newTokensTotal =
+              msg.tokens.input +
+              msg.tokens.output +
+              msg.tokens.reasoning +
+              msg.tokens.cache.read +
+              msg.tokens.cache.write
             if (newTokensTotal > 0) {
               tokensUsedInSession = newTokensTotal
             }
@@ -398,7 +422,9 @@ export async function handleOpencodeSession({
             if (tokensUsedInSession > 0 && usedProviderID && usedModel) {
               if (!modelContextLimit) {
                 try {
-                  const providersResponse = await getClient().provider.list({ query: { directory } })
+                  const providersResponse = await getClient().provider.list({
+                    query: { directory },
+                  })
                   const provider = providersResponse.data?.all?.find((p) => p.id === usedProviderID)
                   const model = provider?.models?.[usedModel]
                   if (model?.limit?.context) {
@@ -410,7 +436,9 @@ export async function handleOpencodeSession({
               }
 
               if (modelContextLimit) {
-                const currentPercentage = Math.floor((tokensUsedInSession / modelContextLimit) * 100)
+                const currentPercentage = Math.floor(
+                  (tokensUsedInSession / modelContextLimit) * 100,
+                )
                 const thresholdCrossed = Math.floor(currentPercentage / 10) * 10
                 if (thresholdCrossed > lastDisplayedContextPercentage && thresholdCrossed >= 10) {
                   lastDisplayedContextPercentage = thresholdCrossed
@@ -431,9 +459,7 @@ export async function handleOpencodeSession({
             continue
           }
 
-          const existingIndex = currentParts.findIndex(
-            (p: Part) => p.id === part.id,
-          )
+          const existingIndex = currentParts.findIndex((p: Part) => p.id === part.id)
           if (existingIndex >= 0) {
             currentParts[existingIndex] = part
           } else {
@@ -468,9 +494,8 @@ export async function handleOpencodeSession({
             const outputTokens = Math.ceil(output.length / 4)
             const LARGE_OUTPUT_THRESHOLD = 3000
             if (outputTokens >= LARGE_OUTPUT_THRESHOLD) {
-              const formattedTokens = outputTokens >= 1000
-                ? `${(outputTokens / 1000).toFixed(1)}k`
-                : String(outputTokens)
+              const formattedTokens =
+                outputTokens >= 1000 ? `${(outputTokens / 1000).toFixed(1)}k` : String(outputTokens)
               const percentageSuffix = (() => {
                 if (!modelContextLimit) {
                   return ''
@@ -519,18 +544,13 @@ export async function handleOpencodeSession({
             const errorData = event.properties.error
             const errorMessage = errorData?.data?.message || 'Unknown error'
             sessionLogger.error(`Sending error to thread: ${errorMessage}`)
-            await sendThreadMessage(
-              thread,
-              `✗ opencode session error: ${errorMessage}`,
-            )
+            await sendThreadMessage(thread, `✗ opencode session error: ${errorMessage}`)
 
             if (originalMessage) {
               try {
                 await originalMessage.reactions.removeAll()
                 await originalMessage.react('❌')
-                voiceLogger.log(
-                  `[REACTION] Added error reaction due to session error`,
-                )
+                voiceLogger.log(`[REACTION] Added error reaction due to session error`)
               } catch (e) {
                 discordLogger.log(`Could not update reaction:`, e)
               }
@@ -579,9 +599,7 @@ export async function handleOpencodeSession({
             continue
           }
 
-          sessionLogger.log(
-            `Permission ${requestID} replied with: ${reply}`,
-          )
+          sessionLogger.log(`Permission ${requestID} replied with: ${reply}`)
 
           const pending = pendingPermissions.get(thread.id)
           if (pending && pending.permission.id === requestID) {
@@ -633,9 +651,7 @@ export async function handleOpencodeSession({
       }
     } catch (e) {
       if (isAbortError(e, abortController.signal)) {
-        sessionLogger.log(
-          'AbortController aborted event handling (normal exit)',
-        )
+        sessionLogger.log('AbortController aborted event handling (normal exit)')
         return
       }
       sessionLogger.error(`Unexpected error in event handling code`, e)
@@ -656,16 +672,12 @@ export async function handleOpencodeSession({
         stopTyping = null
       }
 
-      if (
-        !abortController.signal.aborted ||
-        abortController.signal.reason === 'finished'
-      ) {
-        const sessionDuration = prettyMilliseconds(
-          Date.now() - sessionStartTime,
-        )
+      if (!abortController.signal.aborted || abortController.signal.reason === 'finished') {
+        const sessionDuration = prettyMilliseconds(Date.now() - sessionStartTime)
         const attachCommand = port ? ` ⋅ ${session.id}` : ''
         const modelInfo = usedModel ? ` ⋅ ${usedModel}` : ''
-        const agentInfo = usedAgent && usedAgent.toLowerCase() !== 'build' ? ` ⋅ **${usedAgent}**` : ''
+        const agentInfo =
+          usedAgent && usedAgent.toLowerCase() !== 'build' ? ` ⋅ **${usedAgent}**` : ''
         let contextInfo = ''
 
         try {
@@ -680,8 +692,14 @@ export async function handleOpencodeSession({
           sessionLogger.error('Failed to fetch provider info for context percentage:', e)
         }
 
-        await sendThreadMessage(thread, `_Completed in ${sessionDuration}${contextInfo}_${attachCommand}${modelInfo}${agentInfo}`, { flags: NOTIFY_MESSAGE_FLAGS })
-        sessionLogger.log(`DURATION: Session completed in ${sessionDuration}, port ${port}, model ${usedModel}, tokens ${tokensUsedInSession}`)
+        await sendThreadMessage(
+          thread,
+          `_Completed in ${sessionDuration}${contextInfo}_${attachCommand}${modelInfo}${agentInfo}`,
+          { flags: NOTIFY_MESSAGE_FLAGS },
+        )
+        sessionLogger.log(
+          `DURATION: Session completed in ${sessionDuration}, port ${port}, model ${usedModel}, tokens ${tokensUsedInSession}`,
+        )
 
         // Process queued messages after completion
         const queue = messageQueue.get(thread.id)
@@ -694,7 +712,10 @@ export async function handleOpencodeSession({
           sessionLogger.log(`[QUEUE] Processing queued message from ${nextMessage.username}`)
 
           // Show that queued message is being sent
-          await sendThreadMessage(thread, `» **${nextMessage.username}:** ${nextMessage.prompt.slice(0, 150)}${nextMessage.prompt.length > 150 ? '...' : ''}`)
+          await sendThreadMessage(
+            thread,
+            `» **${nextMessage.username}:** ${nextMessage.prompt.slice(0, 150)}${nextMessage.prompt.length > 150 ? '...' : ''}`,
+          )
 
           // Send the queued message as a new prompt (recursive call)
           // Use setImmediate to avoid blocking and allow this finally to complete
@@ -738,7 +759,14 @@ export async function handleOpencodeSession({
       if (images.length === 0) {
         return prompt
       }
-      sessionLogger.log(`[PROMPT] Sending ${images.length} image(s):`, images.map((img) => ({ mime: img.mime, filename: img.filename, url: img.url.slice(0, 100) })))
+      sessionLogger.log(
+        `[PROMPT] Sending ${images.length} image(s):`,
+        images.map((img) => ({
+          mime: img.mime,
+          filename: img.filename,
+          url: img.url.slice(0, 100),
+        })),
+      )
       const imagePathsList = images.map((img) => `- ${img.filename}: ${img.url}`).join('\n')
       return `${prompt}\n\n**attached images:**\n${imagePathsList}`
     })()
@@ -747,7 +775,8 @@ export async function handleOpencodeSession({
     sessionLogger.log(`[PROMPT] Parts to send:`, parts.length)
 
     // Get model preference: session-level overrides channel-level
-    const modelPreference = getSessionModel(session.id) || (channelId ? getChannelModel(channelId) : undefined)
+    const modelPreference =
+      getSessionModel(session.id) || (channelId ? getChannelModel(channelId) : undefined)
     const modelParam = (() => {
       if (!modelPreference) {
         return undefined
@@ -762,7 +791,8 @@ export async function handleOpencodeSession({
     })()
 
     // Get agent preference: session-level overrides channel-level
-    const agentPreference = getSessionAgent(session.id) || (channelId ? getChannelAgent(channelId) : undefined)
+    const agentPreference =
+      getSessionAgent(session.id) || (channelId ? getChannelAgent(channelId) : undefined)
     if (agentPreference) {
       sessionLogger.log(`[AGENT] Using agent preference: ${agentPreference}`)
     }

package/src/system-message.ts

@@ -2,7 +2,13 @@
 // Creates the system message injected into every OpenCode session,
 // including Discord-specific formatting rules, diff commands, and permissions info.
 
-export function getOpencodeSystemMessage({ sessionId, channelId }: { sessionId: string; channelId?: string }) {
+export function getOpencodeSystemMessage({
+  sessionId,
+  channelId,
+}: {
+  sessionId: string
+  channelId?: string
+}) {
   return `
 The user is reading your messages from inside Discord, via kimaki.xyz
 
@@ -23,15 +29,23 @@ Only users with these Discord permissions can send messages to the bot:
 To upload files to the Discord thread (images, screenshots, long files that would clutter the chat), run:
 
 npx -y kimaki upload-to-discord --session ${sessionId} <file1> [file2] ...
-${channelId ? `
+${
+  channelId
+    ? `
 ## starting new sessions from CLI
 
 To start a new thread/session in this channel programmatically, run:
 
-npx -y kimaki start-session --channel ${channelId} --prompt "your prompt here"
+npx -y kimaki send --channel ${channelId} --prompt "your prompt here"
+
+Use --notify-only to create a notification thread without starting an AI session:
+
+npx -y kimaki send --channel ${channelId} --prompt "User cancelled subscription" --notify-only
 
 This is useful for automation (cron jobs, GitHub webhooks, n8n, etc.)
-` : ''}
+`
+    : ''
+}
 ## showing diffs
 
 IMPORTANT: After editing any files, you MUST execute the critique command using the Bash tool to get a diff URL, then share that URL with the user.