\n \n

Edit Entry

\n \n Text\n \n \n setEditingMemory({ ...editingMemory!, text: e.target.value })\n }\n className=\"w-full p-2 bg-[#fff8dc] border-2 border-[#8B4513] text-[#4b3621] focus:outline-none focus:border-[#654321] font-pixel text-base leading-[1.2em]\"\n rows={3}\n required\n />\n

\n \n Date\n \n \n setEditingMemory({ ...editingMemory!, date: e.target.value })\n }\n className=\"w-full p-1 bg-[#fff8dc] border-2 border-[#8B4513] text-[#4b3621] focus:outline-none focus:border-[#654321] font-pixel text-base\"\n />\n

\n \n Source\n \n \n setEditingMemory({ ...editingMemory!, tags: e.target.value })\n }\n className=\"w-full p-1 bg-[#fff8dc] border-2 border-[#8B4513] text-[#4b3621] focus:outline-none focus:border-[#654321] font-pixel text-base\"\n />\n

\n \n UPDATE\n \n setEditingMemory(null)}\n className=\"px-3 py-1 bg-[#A0522D] text-[#f8f1e0] rounded font-pixel text-base hover:bg-[#8B4513] border-2 border-b-4 border-r-4 border-[#4b3621]\"\n >\n CANCEL\n \n

\n \n editingMemory && handleDeleteMemory(editingMemory.id)\n }\n className=\"px-3 py-1 bg-[#555555] text-white rounded font-pixel text-base hover:bg-[#333333] border-2 border-b-4 border-r-4 border-[#222222]\"\n >\n DELETE\n \n

\n \n

\n );\n\n // Render add form\n const renderAddForm = () => (\n

\n \n

New Entry

\n setNewMemoryText(e.target.value)}\n className=\"w-full p-1 bg-[#fff8dc] border-2 border-[#8B4513] text-[#4b3621] focus:outline-none focus:border-[#654321] font-pixel text-base leading-[1.2em]\"\n rows={2}\n placeholder=\"Enter your note...\"\n required\n />\n

\n \n Date\n \n setNewMemoryDate(e.target.value)}\n className=\"w-full p-1 bg-[#fff8dc] border-2 border-[#8B4513] text-[#4b3621] focus:outline-none focus:border-[#654321] font-pixel text-base\"\n />\n

\n \n Source\n \n setNewMemoryTags(e.target.value)}\n className=\"w-full p-1 bg-[#fff8dc] border-2 border-[#8B4513] text-[#4b3621] focus:outline-none focus:border-[#654321] font-pixel text-base\"\n placeholder=\"Optional source\"\n />\n

\n \n SAVE\n \n setShowAddForm(false)}\n className=\"px-3 py-1 bg-[#A0522D] text-[#f8f1e0] rounded font-pixel text-base hover:bg-[#8B4513] border-2 border-b-4 border-r-4 border-[#4b3621]\"\n >\n CANCEL\n \n

\n \n

\n );\n\n return (\n

\n \n CLOSE\n \n\n

\n Stevens' Notebook\n

\n\n {error && (\n \n ERROR: {error}\n

\n )}\n\n {loading ? (\n

\n ) : (\n

\n \n \n \n \n \n \n \n \n \n \n {paginatedMemories.length === 0 ? (\n \n \n \n ) : (\n paginatedMemories.map((memory) => (\n \n \n \n \n \n \n ))\n )}\n \n

\n DATE\n	\n NOTE\n	\n SOURCE\n	\n EDIT\n
\n \n NO ENTRIES FOUND\n \n
\n \n {formatDate(memory.date)}\n \n	\n \n {memory.text}\n \n	\n \n {memory.tags \|\| \"\"}\n \n	\n handleEditMemory(memory)}\n className=\"mr-1 text-[#8B4513] hover:text-[#654321] font-pixel text-lg\"\n title=\"Edit\"\n >\n ✎\n \n

\n )}\n\n

\n \n PREV\n \n \n NEXT\n \n

\n Page {currentPage} of {totalPages}\n

\n\n

\n {!showAddForm ? (\n setShowAddForm(true)}\n className=\"px-4 py-2 bg-[#8B4513] text-[#f8f1e0] rounded font-pixel text-lg hover:bg-[#654321] shadow-md border-2 border-b-4 border-r-4 border-[#4b3621]\"\n >\n NEW ENTRY\n \n ) : (\n

{renderAddForm()}

\n )}\n

\n\n {/* Edit Modal */}\n {editingMemory && renderEditForm()}\n \n );\n}\n",[186,187,94,92],"4d64935a-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],"/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport React, {\n useState,\n useEffect,\n useCallback,\n useMemo,\n} from \"https://esm.sh/react@18.2.0\";\nimport { type Memory } from \"../../shared/types.ts\";\nimport { ASSETS, SCENE_POSITIONS, SOURCE_TYPES } from \"./assets.ts\";\nimport { NotebookView } from \"./NotebookView.tsx\";\n\nconst API_BASE = \"/api/memories\";\nconst MEMORIES_PER_PAGE = 20; // Increased from 7 to 20 memories per page\n\n// Format date in natural way: \"Wed, April 12\"\nconst formatDate = (dateStr: string) => {\n if (!dateStr) return \"N/A\";\n\n // Parse the date parts manually to avoid timezone issues\n const [year, month, day] = dateStr.split(\"-\").map((num) => parseInt(num, 10));\n\n // Create date with explicit year, month (0-indexed), and day\n const date = new Date(year, month - 1, day);\n\n return date.toLocaleDateString(\"en-US\", {\n weekday: \"short\",\n month: \"long\",\n day: \"numeric\",\n });\n};\n\n// Sort memories by date (ascending)\nconst sortMemoriesByDate = (memories: Memory[]) => {\n return [...memories].sort((a, b) => {\n if (!a.date) return 1;\n if (!b.date) return -1;\n\n // Parse dates manually to avoid timezone issues\n const [aYear, aMonth, aDay] = a.date\n .split(\"-\")\n .map((num) => parseInt(num, 10));\n const [bYear, bMonth, bDay] = b.date\n .split(\"-\")\n .map((num) => parseInt(num, 10));\n\n // Compare year, then month, then day\n if (aYear !== bYear) return aYear - bYear;\n if (aMonth !== bMonth) return aMonth - bMonth;\n return aDay - bDay;\n });\n};\n\n// Scene positions for specific elements on the 512x512 grid\nconst SCENE_ELEMENTS = {\n // Main elements\n DESK: { x: 209, y: 304, width: 84, height: 36 },\n DESK_SITTING: { x: 219, y: 264, width: 64, height: 64 },\n MAILBOX: { x: 61, y: 249, width: 52, height: 79 },\n CALENDAR: { x: 156, y: 226, width: 53, height: 53 },\n TELEGRAM: { x: 308, y: 242, width: 62, height: 80 },\n OUTSIDE: { x: 219, y: 366, width: 64, height: 64 },\n};\n\nexport function App() {\n const [memories, setMemories] = useState([]);\n const [loading, setLoading] = useState(true);\n const [error, setError] = useState(null);\n const [editingMemory, setEditingMemory] = useState(null);\n const [newMemoryText, setNewMemoryText] = useState(\"\");\n const [newMemoryDate, setNewMemoryDate] = useState(\n new Date().toISOString().split(\"T\")[0]\n );\n const [newMemoryTags, setNewMemoryTags] = useState(\"\");\n const [showAddForm, setShowAddForm] = useState(false);\n const [currentPage, setCurrentPage] = useState(1);\n const [avatarUrl, setAvatarUrl] = useState(null);\n const [woodUrl, setWoodUrl] = useState(null);\n const [showNotebook, setShowNotebook] = useState(false);\n const [currentMemoryIndex, setCurrentMemoryIndex] = useState(0);\n const [showCaption, setShowCaption] = useState(false);\n const [isAnimating, setIsAnimating] = useState(false);\n const [cookieAndTeaMode, setCookieAndTeaMode] = useState(false);\n\n // Fetch images from backend instead of blob storage directly\n useEffect(() => {\n // Set default background color in case image doesn't load\n if (document.body) {\n document.body.style.backgroundColor = \"#2D1700\"; // Dark brown leather color\n }\n\n // Fetch avatar image\n fetch(\"/api/images/stevens.jpg\")\n .then((response) => {\n if (response.ok) return response.blob();\n throw new Error(\"Failed to load avatar image\");\n })\n .then((imageBlob) => {\n const url = URL.createObjectURL(imageBlob);\n setAvatarUrl(url);\n })\n .catch((err) => {\n console.error(\"Failed to load avatar:\", err);\n });\n\n // Fetch wood background\n fetch(\"/api/images/wood.jpg\")\n .then((response) => {\n if (response.ok) return response.blob();\n throw new Error(\"Failed to load wood background\");\n })\n .then((imageBlob) => {\n const url = URL.createObjectURL(imageBlob);\n setWoodUrl(url);\n\n // Apply wood background to body\n if (document.body) {\n document.body.style.backgroundImage = `url(${url})`;\n }\n })\n .catch((err) => {\n console.error(\"Failed to load background:\", err);\n });\n\n // Cleanup blob URLs on unmount\n return () => {\n if (avatarUrl) URL.revokeObjectURL(avatarUrl);\n if (woodUrl) URL.revokeObjectURL(woodUrl);\n };\n }, []);\n\n const fetchMemories = useCallback(async () => {\n setLoading(true);\n setError(null);\n try {\n const response = await fetch(API_BASE);\n if (!response.ok) {\n throw new Error(`HTTP error! status: ${response.status}`);\n }\n const data = await response.json();\n\n // Change the sorting function to show memories in chronological order\n const sortedMemories = [...data].sort((a, b) => {\n const dateA = a.createdDate || 0;\n const dateB = b.createdDate || 0;\n return dateA - dateB; // Oldest first, chronological order\n });\n\n setMemories(sortedMemories);\n\n // Set current index to the last memory (newest one)\n if (sortedMemories.length > 0) {\n setCurrentMemoryIndex(sortedMemories.length - 1);\n setShowCaption(true);\n }\n } catch (e) {\n console.error(\"Failed to fetch memories:\", e);\n setError(e.message || \"Failed to fetch memories.\");\n } finally {\n setLoading(false);\n }\n }, []);\n\n useEffect(() => {\n fetchMemories();\n }, [fetchMemories]);\n\n const handleAddMemory = async (e: React.FormEvent) => {\n e.preventDefault();\n if (!newMemoryText.trim()) return;\n\n const memoryData: Omit = {\n text: newMemoryText,\n date: newMemoryDate,\n tags: newMemoryTags || null,\n };\n\n try {\n const response = await fetch(API_BASE, {\n method: \"POST\",\n headers: { \"Content-Type\": \"application/json\" },\n body: JSON.stringify(memoryData),\n });\n if (!response.ok) {\n throw new Error(`HTTP error! status: ${response.status}`);\n }\n setNewMemoryText(\"\");\n setNewMemoryDate(new Date().toISOString().split(\"T\")[0]);\n setNewMemoryTags(\"\");\n setShowAddForm(false);\n await fetchMemories();\n } catch (e) {\n console.error(\"Failed to add memory:\", e);\n setError(e.message || \"Failed to add memory.\");\n }\n };\n\n const handleDeleteMemory = async (id: string) => {\n if (!confirm(\"Are you sure you want to delete this memory?\")) return;\n\n try {\n const response = await fetch(`${API_BASE}/${id}`, {\n method: \"DELETE\",\n });\n if (!response.ok) {\n throw new Error(`HTTP error! status: ${response.status}`);\n }\n await fetchMemories();\n } catch (e) {\n console.error(\"Failed to delete memory:\", e);\n setError(e.message || \"Failed to delete memory.\");\n }\n };\n\n const handleEditMemory = (memory: Memory) => {\n setEditingMemory(memory);\n };\n\n const handleCancelEdit = () => {\n setEditingMemory(null);\n };\n\n const handleUpdateMemory = async (e: React.FormEvent) => {\n e.preventDefault();\n if (!editingMemory || !editingMemory.text.trim()) return;\n\n const updatedFields: Partial> = {\n text: editingMemory.text,\n date: editingMemory.date,\n tags: editingMemory.tags,\n };\n\n try {\n const response = await fetch(`${API_BASE}/${editingMemory.id}`, {\n method: \"PUT\",\n headers: { \"Content-Type\": \"application/json\" },\n body: JSON.stringify(updatedFields),\n });\n if (!response.ok) {\n throw new Error(`HTTP error! status: ${response.status}`);\n }\n setEditingMemory(null);\n await fetchMemories();\n } catch (e) {\n console.error(\"Failed to update memory:\", e);\n setError(e.message || \"Failed to update memory.\");\n }\n };\n\n // Sort and paginate memories\n const sortedMemories = useMemo(\n () => sortMemoriesByDate(memories),\n [memories]\n );\n\n const totalPages = Math.max(\n 1,\n Math.ceil(sortedMemories.length / MEMORIES_PER_PAGE)\n );\n\n const paginatedMemories = useMemo(() => {\n const startIndex = (currentPage - 1) * MEMORIES_PER_PAGE;\n return sortedMemories.slice(startIndex, startIndex + MEMORIES_PER_PAGE);\n }, [sortedMemories, currentPage]);\n\n const goToNextPage = () => {\n if (currentPage < totalPages) {\n setCurrentPage(currentPage + 1);\n }\n };\n\n const goToPrevPage = () => {\n if (currentPage > 1) {\n setCurrentPage(currentPage - 1);\n }\n };\n\n // Animation delay to keep Stevens in place briefly\n const handleAnimationEnd = () => {\n setIsAnimating(false);\n };\n\n // Memory navigation handlers\n const goToNextMemory = () => {\n if (currentMemoryIndex < memories.length - 1) {\n // Reset cookie and tea mode on navigation\n setCookieAndTeaMode(false);\n\n // Trigger animation transition\n setIsAnimating(true);\n\n // Change memory after a brief delay\n setCurrentMemoryIndex(currentMemoryIndex + 1);\n setShowCaption(true);\n }\n };\n\n const goToPrevMemory = () => {\n if (currentMemoryIndex > 0) {\n // Reset cookie and tea mode on navigation\n setCookieAndTeaMode(false);\n\n // Trigger animation transition\n setIsAnimating(true);\n\n // Change memory after a brief delay\n setCurrentMemoryIndex(currentMemoryIndex - 1);\n setShowCaption(true);\n }\n };\n\n // Add keyboard navigation\n useEffect(() => {\n const handleKeyDown = (e: KeyboardEvent) => {\n if (showNotebook) return; // Don't navigate when notebook is open\n\n if (e.key === \"ArrowLeft\") {\n goToPrevMemory();\n } else if (e.key === \"ArrowRight\") {\n goToNextMemory();\n }\n };\n\n window.addEventListener(\"keydown\", handleKeyDown);\n\n return () => {\n window.removeEventListener(\"keydown\", handleKeyDown);\n };\n }, [currentMemoryIndex, memories.length, showNotebook]);\n\n // Generate caption based on memory source and content\n const generateCaption = (memory: Memory | undefined) => {\n if (cookieAndTeaMode) {\n return \"Ah, a lovely cup of tea and a cookie. Most refreshing indeed.\";\n }\n\n if (!memory) return \"No memories to display\";\n\n const source = memory.createdBy || \"unknown\";\n const sourceText =\n source === SOURCE_TYPES.WEATHER\n ? \"Hmm, I should check the weather today.\"\n : source === SOURCE_TYPES.MAIL\n ? \"Ah! A new letter has arrived.\"\n : source === SOURCE_TYPES.TELEGRAM\n ? \"A telegram! I must read it right away.\"\n : source === SOURCE_TYPES.CALENDAR\n ? \"I should check my calendar.\"\n : \"I must make note of this.\";\n\n return `${sourceText} Note for ${memory.date}: ${memory.text}`;\n };\n\n // Handle cookie and tea button\n const handleServeCookieAndTea = () => {\n setCookieAndTeaMode(true);\n };\n\n // Get Stevens' position and appearance based on memory source\n const getStevensState = (memory: Memory | undefined) => {\n // Override with cookie and tea mode\n if (cookieAndTeaMode) {\n return {\n position: SCENE_ELEMENTS.DESK_SITTING,\n image: ASSETS.STEVENS_FRONT,\n highlightElement: SCENE_ELEMENTS.DESK,\n animationClass: \"no-animation\",\n };\n }\n\n if (!memory) {\n // Default state: sitting at desk\n return {\n position: SCENE_ELEMENTS.DESK_SITTING,\n image: ASSETS.STEVENS_FRONT,\n highlightElement: null,\n animationClass: \"no-animation\",\n };\n }\n\n const source = memory.createdBy || \"unknown\";\n\n switch (source) {\n case SOURCE_TYPES.MAIL:\n return {\n position: {\n x: SCENE_ELEMENTS.MAILBOX.x + 20,\n y: SCENE_ELEMENTS.MAILBOX.y - 20,\n },\n image: ASSETS.STEVENS_BACK,\n highlightElement: SCENE_ELEMENTS.MAILBOX,\n animationClass: \"walk-to-mailbox\",\n };\n\n case SOURCE_TYPES.CALENDAR:\n return {\n position: {\n x: SCENE_ELEMENTS.CALENDAR.x + 10,\n y: SCENE_ELEMENTS.CALENDAR.y + 30,\n },\n image: ASSETS.STEVENS_BACK,\n highlightElement: SCENE_ELEMENTS.CALENDAR,\n animationClass: \"walk-to-calendar\",\n };\n\n case SOURCE_TYPES.TELEGRAM:\n return {\n position: {\n x: SCENE_ELEMENTS.TELEGRAM.x - 20,\n y: SCENE_ELEMENTS.TELEGRAM.y + 10,\n },\n image: ASSETS.STEVENS_BACK,\n highlightElement: SCENE_ELEMENTS.TELEGRAM,\n animationClass: \"walk-to-telegram\",\n };\n\n case SOURCE_TYPES.WEATHER:\n return {\n position: SCENE_ELEMENTS.OUTSIDE,\n image: ASSETS.STEVENS_FRONT,\n highlightElement: null,\n animationClass: \"walk-to-outside\",\n };\n\n default:\n return {\n position: SCENE_ELEMENTS.DESK_SITTING,\n image: ASSETS.STEVENS_FRONT,\n highlightElement: SCENE_ELEMENTS.DESK,\n animationClass: \"walk-to-desk\",\n };\n }\n };\n\n // Current memory\n const currentMemory = memories[currentMemoryIndex];\n const caption = generateCaption(currentMemory);\n const stevensState = getStevensState(currentMemory);\n\n // Scale factor for world coordinates (1/2 original size)\n const scaleFactor = 0.5;\n const scalePosition = (pos: { x: number; y: number }) => ({\n x: pos.x * scaleFactor,\n y: pos.y * scaleFactor,\n });\n\n // Render edit form modal\n const renderEditForm = () => (\n

\n \n

Edit Entry

\n \n DELETE\n \n

\n \n

\n );\n\n // Render add form\n const renderAddForm = () => (\n

\n \n

New Entry

\n \n

\n );\n\n return (\n

\n \n\n {/* Main game container with fixed dimensions */}\n

\n {/* Game world */}\n \n {/* Notebook clickable area (transparent) */}\n setShowNotebook(true)}\n title=\"Click to open Stevens' notebook\"\n >

\n\n {/* Highlight for interactive elements */}\n {stevensState.highlightElement && (\n

\n )}\n\n {/* Stevens character */}\n \n\n {/* Cookie and teacup for easter egg */}\n {cookieAndTeaMode && (\n <>\n \n \n \n )}\n \n\n {/* Caption area */}\n

\n {showCaption && (currentMemory || cookieAndTeaMode) ? (\n

{caption}

\n ) : (\n

\n Use arrows or buttons to navigate through Stevens' memories\n

\n )}\n

\n\n {/* Timeline controls */}\n

\n {loading\n ? \"Loading memories...\"\n : `Memory ${currentMemoryIndex + 1} of ${memories.length}`}\n

\n\n

\n \n ◀ PREV\n \n\n = memories.length - 1}\n className={`px-4 py-2 rounded-lg border-2 border-b-4 border-r-4 ${\n currentMemoryIndex >= memories.length - 1\n ? \"bg-gray-600 border-gray-700 opacity-50 cursor-not-allowed\"\n : \"bg-[#8B4513] border-[#654321] hover:bg-[#A0522D]\"\n }`}\n aria-label=\"Next memory\"\n >\n NEXT ▶\n \n

\n\n {error && (\n

\n {error}\n

\n )}\n\n {/* Timeline visualization and cookie/tea button */}\n

\n Source:{\" \"}\n {cookieAndTeaMode\n ? \"Tea Time\"\n : currentMemory?.createdBy || \"None\"}\n

\n Use ◀ ▶ arrow keys or buttons to navigate\n

\n\n {/* Easter egg button */}\n \n 🍪☕ Serve Cookie & Tea\n \n

\n\n {/* Notebook View Modal */}\n {showNotebook && (\n setShowNotebook(false)}\n avatarUrl={avatarUrl}\n />\n )}\n \n \n );\n}\n",[186,187,94,75],"pulls","branches","forks","envVars",{"_88":26,"_211":26},"connections",{"_88":26,"_210":26},"unfilledCount","emptyCount",{"_33":520,"_13":505,"_73":500,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":521,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":522,"_17":523},{"_33":513,"_13":514,"_73":515,"_43":500,"_35":36,"_180":40,"_181":24,"_37":516,"_39":26,"_76":325,"_183":517,"_78":30,"_184":-5,"_260":518,"_17":519},{"_33":506,"_13":507,"_73":508,"_43":500,"_35":36,"_180":40,"_181":24,"_37":509,"_39":26,"_76":77,"_183":510,"_78":30,"_184":-5,"_260":511,"_17":512},{"_33":497,"_13":498,"_73":499,"_43":500,"_35":36,"_180":40,"_181":24,"_37":501,"_39":26,"_76":77,"_183":502,"_78":30,"_184":-5,"_260":503,"_17":504},{"_33":493,"_13":186,"_73":383,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":494,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":495,"_17":496},{"_33":489,"_13":463,"_73":458,"_43":383,"_35":36,"_180":40,"_181":24,"_37":490,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":491,"_17":492},{"_33":485,"_13":478,"_73":473,"_43":458,"_35":36,"_180":40,"_181":24,"_37":486,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":487,"_17":488},{"_33":479,"_13":278,"_73":480,"_43":473,"_35":36,"_180":40,"_181":24,"_37":481,"_39":26,"_76":21,"_183":482,"_78":30,"_184":-5,"_260":483,"_17":484},{"_33":470,"_13":471,"_73":472,"_43":473,"_35":36,"_180":40,"_181":24,"_37":474,"_39":26,"_76":77,"_183":475,"_78":30,"_184":-5,"_260":476,"_17":477},{"_33":464,"_13":278,"_73":465,"_43":458,"_35":36,"_180":40,"_181":24,"_37":466,"_39":26,"_76":21,"_183":467,"_78":30,"_184":-5,"_260":468,"_17":469},{"_33":455,"_13":456,"_73":457,"_43":458,"_35":36,"_180":40,"_181":24,"_37":459,"_39":26,"_76":317,"_183":460,"_78":30,"_184":-5,"_260":461,"_17":462},{"_33":451,"_13":187,"_73":179,"_43":383,"_35":36,"_180":40,"_181":24,"_37":452,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":453,"_17":454},{"_33":178,"_13":94,"_73":93,"_43":179,"_35":36,"_180":40,"_181":24,"_37":448,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":449,"_17":450},{"_33":199,"_13":75,"_73":74,"_43":93,"_35":36,"_180":40,"_181":24,"_37":445,"_39":26,"_76":77,"_183":201,"_78":30,"_184":-5,"_260":446,"_17":447},{"_33":195,"_13":92,"_73":91,"_43":93,"_35":36,"_180":40,"_181":24,"_37":442,"_39":26,"_76":77,"_183":197,"_78":30,"_184":-5,"_260":443,"_17":444},{"_33":191,"_13":87,"_73":86,"_43":93,"_35":36,"_180":40,"_181":24,"_37":439,"_39":26,"_76":77,"_183":193,"_78":30,"_184":-5,"_260":440,"_17":441},{"_33":433,"_13":278,"_73":434,"_43":179,"_35":36,"_180":40,"_181":24,"_37":435,"_39":26,"_76":21,"_183":436,"_78":30,"_184":-5,"_260":437,"_17":438},{"_33":426,"_13":427,"_73":428,"_43":179,"_35":36,"_180":40,"_181":24,"_37":429,"_39":26,"_76":21,"_183":430,"_78":30,"_184":-5,"_260":431,"_17":432},{"_33":419,"_13":420,"_73":421,"_43":179,"_35":36,"_180":40,"_181":24,"_37":422,"_39":26,"_76":21,"_183":423,"_78":30,"_184":-5,"_260":424,"_17":425},{"_33":412,"_13":413,"_73":414,"_43":179,"_35":36,"_180":40,"_181":24,"_37":415,"_39":26,"_76":77,"_183":416,"_78":30,"_184":-5,"_260":417,"_17":418},{"_33":408,"_13":299,"_73":391,"_43":383,"_35":36,"_180":40,"_181":24,"_37":409,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":410,"_17":411},{"_33":402,"_13":278,"_73":403,"_43":391,"_35":36,"_180":40,"_181":24,"_37":404,"_39":26,"_76":21,"_183":405,"_78":30,"_184":-5,"_260":406,"_17":407},{"_33":396,"_13":292,"_73":397,"_43":391,"_35":36,"_180":40,"_181":24,"_37":398,"_39":26,"_76":77,"_183":399,"_78":30,"_184":-5,"_260":400,"_17":401},{"_33":388,"_13":389,"_73":390,"_43":391,"_35":36,"_180":40,"_181":24,"_37":392,"_39":26,"_76":77,"_183":393,"_78":30,"_184":-5,"_260":394,"_17":395},{"_33":381,"_13":278,"_73":382,"_43":383,"_35":36,"_180":40,"_181":24,"_37":384,"_39":26,"_76":21,"_183":385,"_78":30,"_184":-5,"_260":386,"_17":387},{"_33":377,"_13":355,"_73":350,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":378,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":379,"_17":380},{"_33":370,"_13":371,"_73":372,"_43":350,"_35":36,"_180":40,"_181":24,"_37":373,"_39":26,"_76":77,"_183":374,"_78":30,"_184":-5,"_260":375,"_17":376},{"_33":363,"_13":364,"_73":365,"_43":350,"_35":36,"_180":40,"_181":24,"_37":366,"_39":26,"_76":77,"_183":367,"_78":30,"_184":-5,"_260":368,"_17":369},{"_33":356,"_13":357,"_73":358,"_43":350,"_35":36,"_180":40,"_181":24,"_37":359,"_39":26,"_76":77,"_183":360,"_78":30,"_184":-5,"_260":361,"_17":362},{"_33":347,"_13":348,"_73":349,"_43":350,"_35":36,"_180":40,"_181":24,"_37":351,"_39":26,"_76":77,"_183":352,"_78":30,"_184":-5,"_260":353,"_17":354},{"_33":343,"_13":313,"_73":307,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":344,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":345,"_17":346},{"_33":336,"_13":337,"_73":338,"_43":307,"_35":36,"_180":40,"_181":24,"_37":339,"_39":26,"_76":325,"_183":340,"_78":30,"_184":-5,"_260":341,"_17":342},{"_33":329,"_13":330,"_73":331,"_43":307,"_35":36,"_180":40,"_181":24,"_37":332,"_39":26,"_76":325,"_183":333,"_78":30,"_184":-5,"_260":334,"_17":335},{"_33":321,"_13":322,"_73":323,"_43":307,"_35":36,"_180":40,"_181":24,"_37":324,"_39":26,"_76":325,"_183":326,"_78":30,"_184":-5,"_260":327,"_17":328},{"_33":314,"_13":315,"_73":106,"_43":307,"_35":36,"_180":40,"_181":24,"_37":316,"_39":26,"_76":317,"_183":318,"_78":30,"_184":-5,"_260":319,"_17":320},{"_33":304,"_13":305,"_73":306,"_43":307,"_35":36,"_180":40,"_181":24,"_37":308,"_39":26,"_76":309,"_183":310,"_78":30,"_184":-5,"_260":311,"_17":312},{"_33":300,"_13":299,"_73":294,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":301,"_39":26,"_76":95,"_183":-5,"_78":30,"_184":-5,"_260":302,"_17":303},{"_33":291,"_13":292,"_73":293,"_43":294,"_35":36,"_180":40,"_181":24,"_37":295,"_39":26,"_76":77,"_183":296,"_78":30,"_184":-5,"_260":297,"_17":298},{"_33":284,"_13":285,"_73":286,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":287,"_39":26,"_76":21,"_183":288,"_78":30,"_184":-5,"_260":289,"_17":290},{"_33":277,"_13":278,"_73":279,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":280,"_39":26,"_76":21,"_183":281,"_78":30,"_184":-5,"_260":282,"_17":283},{"_33":270,"_13":271,"_73":272,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":273,"_39":26,"_76":77,"_183":274,"_78":30,"_184":-5,"_260":275,"_17":276},{"_33":263,"_13":264,"_73":265,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":266,"_39":26,"_76":77,"_183":267,"_78":30,"_184":-5,"_260":268,"_17":269},{"_33":255,"_13":256,"_73":257,"_43":-5,"_35":36,"_180":40,"_181":24,"_37":258,"_39":26,"_76":77,"_183":259,"_78":30,"_184":-5,"_260":261,"_17":262},"4d65e2d2-3d2c-11f0-bd13-76b3cceeab13","populateDemo.ts","bd64a873-7c7e-43ac-8299-29b0e0ef3559",["D",1748592093410],"import { nanoid } from \"https://esm.sh/nanoid@5.0.5\";\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\n\nconst TABLE_NAME = \"memories_demo\";\n\n// Create the memories_demo table\nasync function createMemoriesDemoTable() {\n try {\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS ${TABLE_NAME} (\n id TEXT PRIMARY KEY,\n date TEXT,\n text TEXT NOT NULL,\n createdBy TEXT,\n createdDate NUMBER,\n tags TEXT\n )\n `);\n console.log(`Created table: ${TABLE_NAME}`);\n } catch (error) {\n console.error(\"Error creating memories_demo table:\", error);\n throw error;\n }\n}\n\n// Create a fake memory with proper ID and timestamps\nfunction createMemory(date, text, createdBy, tags, createdDateOffset = 0) {\n const id = nanoid(10);\n // Base date is April 5, 2025\n const baseDate = new Date(\"2025-04-05\").getTime();\n // Add random offset to createdDate (within reasonable bounds)\n const createdDate = baseDate + createdDateOffset;\n\n return {\n id,\n date, // Can be null for undated memories\n text,\n createdBy,\n createdDate,\n tags,\n };\n}\n\n// Seed data for the Doe family\nconst demoMemories = [\n // Weather forecasts\n createMemory(\n \"2025-04-11\",\n \"weather forecast: High of 72, low of 54, sunny morning giving way to scattered clouds in the afternoon.\",\n \"weather\",\n \"weather\",\n -518400000,\n ),\n createMemory(\n \"2025-04-12\",\n \"weather forecast: High of 68, low of 51, partly cloudy with a 30% chance of light showers in the evening.\",\n \"weather\",\n \"weather\",\n -432000000,\n ),\n createMemory(\n \"2025-04-13\",\n \"weather forecast: High of 65, low of 48, overcast with intermittent drizzle throughout the day.\",\n \"weather\",\n \"weather\",\n -345600000,\n ),\n createMemory(\n \"2025-04-14\",\n \"weather forecast: High of 62, low of 46, morning fog clearing to reveal sunny skies by midday.\",\n \"weather\",\n \"weather\",\n -259200000,\n ),\n createMemory(\n \"2025-04-15\",\n \"weather forecast: High of 70, low of 53, clear and mild with light breezes from the southwest.\",\n \"weather\",\n \"weather\",\n -172800000,\n ),\n createMemory(\n \"2025-04-16\",\n \"weather forecast: High of 75, low of 57, warm and mostly sunny with increasing humidity.\",\n \"weather\",\n \"weather\",\n -86400000,\n ),\n createMemory(\n \"2025-04-17\",\n \"weather forecast: High of 77, low of 60, hot and humid with potential thunderstorms developing late afternoon.\",\n \"weather\",\n \"weather\",\n 0,\n ),\n\n // Calendar events\n createMemory(\n \"2025-04-12\",\n \"Lucas's soccer practice from 10:00 am - 11:30 am at Oakridge Community Fields\",\n \"calendar\",\n \"calendar\",\n -518400000,\n ),\n createMemory(\n \"2025-04-14\",\n \"Emma's science club meeting from 3:30 pm - 5:00 pm at Lincoln Middle School\",\n \"calendar\",\n \"calendar\",\n -432000000,\n ),\n createMemory(\n \"2025-04-15\",\n \"Parent-teacher conference from 4:00 pm - 4:30 pm at Lincoln Middle School\",\n \"calendar\",\n \"calendar\",\n -345600000,\n ),\n createMemory(\n \"2025-04-16\",\n \"John's dental checkup from 2:15 pm - 3:00 pm at Brightsmile Dental Clinic\",\n \"calendar\",\n \"calendar\",\n -259200000,\n ),\n createMemory(\n \"2025-04-17\",\n \"Jane's research presentation from 1:00 pm - 2:30 pm at University Biology Department\",\n \"calendar\",\n \"calendar\",\n -172800000,\n ),\n createMemory(\n \"2025-04-19\",\n \"Family dinner reservation from 6:30 pm - 8:30 pm at Riverside Italian Bistro\",\n \"calendar\",\n \"calendar\",\n -86400000,\n ),\n createMemory(\n \"2025-04-21\",\n \"Emma's piano recital from 5:00 pm - 7:00 pm at Community Arts Center\",\n \"calendar\",\n \"calendar\",\n 0,\n ),\n\n // Fun facts\n createMemory(\n \"2025-04-11\",\n \"fun fact: Border collies can learn over 1,000 words and recognize objects by name, rivaling the linguistic capabilities of a 2-year-old human child.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n -518400000,\n ),\n createMemory(\n \"2025-04-12\",\n \"fun fact: Monstera plants develop their characteristic holes through a process called fenestration, which helps them resist damage from heavy rainfall and allows light to reach lower leaves.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n -432000000,\n ),\n createMemory(\n \"2025-04-13\",\n \"fun fact: Elephant calves can stand within 20 minutes of birth and can walk within 1 hour, an evolutionary adaptation that helps them keep up with the constantly moving herd.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n -345600000,\n ),\n createMemory(\n \"2025-04-14\",\n \"fun fact: At two years old, children typically have a vocabulary of 50-200 words and are beginning to form simple two-word sentences like 'more juice' or 'daddy home'.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n -259200000,\n ),\n createMemory(\n \"2025-04-15\",\n \"fun fact: Orchids have evolved specialized relationships with specific pollinators; some species emit scents that mimic female insects to attract males for pollination.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n -172800000,\n ),\n createMemory(\n \"2025-04-16\",\n \"fun fact: Penguins propose to their mates with a pebble, and if accepted, the pebble is used in their nest. They typically mate for life.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n -86400000,\n ),\n createMemory(\n \"2025-04-17\",\n \"fun fact: By age two, toddlers can recognize familiar melodies and may begin to sing along with simple songs, demonstrating early musical development.\",\n \"fun_fact_generator\",\n \"fun_fact\",\n 0,\n ),\n\n // USPS package tracking\n createMemory(\n \"2025-04-12\",\n \"expecting package: [{\\\"sender\\\":\\\"BOOK DEPOSITORY\\\",\\\"status\\\":\\\"awaiting from sender\\\"}]\",\n \"usps\",\n \"mail\",\n -518400000,\n ),\n createMemory(\n \"2025-04-13\",\n \"expecting package: [{\\\"sender\\\":\\\"BOOK DEPOSITORY\\\",\\\"status\\\":\\\"expected today\\\"},{\\\"sender\\\":\\\"AMAZON\\\",\\\"status\\\":\\\"awaiting from sender\\\"}]\",\n \"usps\",\n \"mail\",\n -432000000,\n ),\n createMemory(\n \"2025-04-14\",\n \"expecting package: [{\\\"sender\\\":\\\"AMAZON\\\",\\\"status\\\":\\\"expected 1-2 days\\\"},{\\\"sender\\\":\\\"UNIVERSITY BOOKSTORE\\\",\\\"status\\\":\\\"awaiting from sender\\\"}]\",\n \"usps\",\n \"mail\",\n -345600000,\n ),\n createMemory(\n \"2025-04-15\",\n \"expecting package: [{\\\"sender\\\":\\\"UNIVERSITY BOOKSTORE\\\",\\\"status\\\":\\\"expected today\\\"},{\\\"sender\\\":\\\"GARDENING SUPPLY CO\\\",\\\"status\\\":\\\"awaiting from sender\\\"}]\",\n \"usps\",\n \"mail\",\n -259200000,\n ),\n createMemory(\n \"2025-04-16\",\n \"expecting package: [{\\\"sender\\\":\\\"GARDENING SUPPLY CO\\\",\\\"status\\\":\\\"expected 1-2 days\\\"},{\\\"sender\\\":\\\"SOCCER EQUIPMENT PLUS\\\",\\\"status\\\":\\\"awaiting from sender\\\"}]\",\n \"usps\",\n \"mail\",\n -172800000,\n ),\n\n // Mail\n createMemory(\n \"2025-04-12\",\n \"expecting mail: [{\\\"sender\\\":\\\"Lincoln Middle School\\\",\\\"recipient\\\":\\\"John\\\",\\\"type\\\":\\\"transactional\\\",\\\"notes\\\":\\\"Science Fair Information\\\"}]\",\n \"usps\",\n \"mail\",\n -518400000,\n ),\n createMemory(\n \"2025-04-13\",\n \"expecting mail: [{\\\"sender\\\":\\\"University Biology Department\\\",\\\"recipient\\\":\\\"Jane\\\",\\\"type\\\":\\\"transactional\\\",\\\"notes\\\":\\\"Research Grant Approval\\\"}]\",\n \"usps\",\n \"mail\",\n -432000000,\n ),\n createMemory(\n \"2025-04-15\",\n \"expecting mail: [{\\\"sender\\\":\\\"City Water Utility\\\",\\\"recipient\\\":\\\"John\\\",\\\"type\\\":\\\"transactional\\\"}]\",\n \"usps\",\n \"mail\",\n -259200000,\n ),\n\n // Telegram messages\n createMemory(\n \"2025-04-11\",\n \"Jane prefers Earl Grey tea in the morning.\",\n \"telegram\",\n \"telegram\",\n -518400000,\n ),\n createMemory(\n \"2025-04-12\",\n \"John's parents visiting for dinner on Sunday, April 20th. They're vegetarians.\",\n \"telegram\",\n \"telegram\",\n -432000000,\n ),\n createMemory(\n \"2025-04-13\",\n \"Lucas allergic to peanuts - mild reaction only, not anaphylactic.\",\n \"telegram\",\n \"telegram\",\n -345600000,\n ),\n createMemory(\n \"2025-04-14\",\n \"Emma's birthday party theme will be astronomy/space.\",\n \"telegram\",\n \"telegram\",\n -259200000,\n ),\n createMemory(\n \"2025-04-15\",\n \"Jane working from home on Mondays and Wednesdays.\",\n \"telegram\",\n \"telegram\",\n -172800000,\n ),\n createMemory(\n \"2025-04-16\",\n \"John has a deadline for software project on April 22nd.\",\n \"telegram\",\n \"telegram\",\n -86400000,\n ),\n createMemory(\n \"2025-04-17\",\n \"Both children need to bring packed lunches for school field trip on April 18th.\",\n \"telegram\",\n \"telegram\",\n 0,\n ),\n\n // Undated memories (general family information)\n createMemory(\n null,\n \"John Doe works as a software engineer at TechCorp.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"Jane Doe works as a biology professor at State University.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"Emma Doe is 10 years old, interested in astronomy and piano.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"Lucas Doe is 7 years old, interested in soccer and dinosaurs.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"Family grocery shopping usually at Whole Foods or Trader Joe's.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"John's parents are Robert and Martha Doe, living in Chicago.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"Jane's sister Sarah lives nearby and sometimes picks up the children from school.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"Emma's birthday is June 15, Lucas's birthday is September 22.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n createMemory(\n null,\n \"John and Jane's anniversary is October 8.\",\n \"telegram\",\n \"telegram\",\n -604800000,\n ),\n];\n\n// Insert memories into the database\nasync function insertDemoMemories() {\n try {\n // Clear existing data if any\n await sqlite.execute(`DELETE FROM ${TABLE_NAME}`);\n\n let insertedCount = 0;\n\n for (const memory of demoMemories) {\n await sqlite.execute(\n `INSERT INTO ${TABLE_NAME} (id, date, text, createdBy, createdDate, tags)\n VALUES (?, ?, ?, ?, ?, ?)`,\n [\n memory.id,\n memory.date,\n memory.text,\n memory.createdBy,\n memory.createdDate,\n memory.tags,\n ],\n );\n insertedCount++;\n }\n\n console.log(`Successfully inserted ${insertedCount} demo memories.`);\n return `Successfully inserted ${insertedCount} demo memories.`;\n } catch (error) {\n console.error(\"Error inserting demo memories:\", error);\n throw error;\n }\n}\n\n// Main function to populate demo data\nexport default async function populateDemo() {\n try {\n // Create the table\n await createMemoriesDemoTable();\n\n // Insert demo memories\n await insertDemoMemories();\n\n console.log(\"Demo database successfully populated!\");\n return \"Demo database successfully populated!\";\n } catch (error) {\n console.error(\"Error populating demo database:\", error);\n throw error;\n }\n}\n\n// Uncomment to run directly when testing\npopulateDemo();","children",[],[256],"4d65ddfa-3d2c-11f0-bd13-76b3cceeab13","memoryUtils.ts","63665fcb-2ce1-47e4-bf43-26e18c7cb5b4",["D",1748592093410],"import { DateTime } from \"https://esm.sh/luxon@3.4.4\";\n\n/**\n * Retrieves all memories from the database\n * @param includeDate Whether to include date-specific memories or not\n * @param startDate Optional start date to filter memories from (ISO format)\n * @returns Array of memory objects\n */\nexport async function getAllMemories(includeDate = true, startDate = null) {\n try {\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n // Get memories with dates if requested\n let memories = [];\n if (includeDate) {\n const query = startDate\n ? {\n sql: `SELECT id, date(date) as date, text FROM memories WHERE date >= :startDate AND date <= date(:startDate, '+7 days') ORDER BY date ASC`,\n args: { startDate },\n }\n : `SELECT id, date(date) as date, text FROM memories WHERE date IS NOT NULL ORDER BY date ASC`;\n\n const dateMemories = await sqlite.execute(query);\n memories = [...dateMemories.rows];\n }\n\n // Always get dateless memories\n const datelessMemories = await sqlite.execute(\n `SELECT id, date, text FROM memories WHERE date IS NULL`\n );\n\n return [...memories, ...datelessMemories.rows];\n } catch (error) {\n console.error(\"Error retrieving memories:\", error);\n return [];\n }\n}\n\n/**\n * Gets memories relevant to today and future dates\n * @returns Array of memory objects\n */\nexport async function getRelevantMemories() {\n try {\n // Get today's date in US Eastern Time\n const today = DateTime.now().setZone(\"America/New_York\").startOf(\"day\");\n const todayFormatted = today.toFormat(\"yyyy-MM-dd\");\n\n return await getAllMemories(true, todayFormatted);\n } catch (error) {\n console.error(\"Error getting relevant memories:\", error);\n return [];\n }\n}\n\n/**\n * Formats memories into a string for the system prompt\n * @param memories Array of memory objects\n * @returns Formatted string of memories\n */\nexport function formatMemoriesForPrompt(memories) {\n if (!memories || memories.length === 0) {\n return \"No stored memories are available.\";\n }\n\n // Format dated memories\n const datedMemories = memories\n .filter((memory) => memory.date)\n .map((memory) => {\n const date = DateTime.fromISO(memory.date);\n return `- ${date.toFormat(\"yyyy-MM-dd\")} [ID: ${memory.id}]: ${\n memory.text\n }`;\n });\n\n // Format dateless memories\n const datelessMemories = memories\n .filter((memory) => !memory.date)\n .map((memory) => `- [ID: ${memory.id}]: ${memory.text}`);\n\n let result = \"\";\n\n if (datedMemories.length > 0) {\n result += \"Dated memories:\\n\" + datedMemories.join(\"\\n\") + \"\\n\\n\";\n }\n\n if (datelessMemories.length > 0) {\n result += \"General memories:\\n\" + datelessMemories.join(\"\\n\");\n }\n\n return result;\n}\n",[],[264],"4d65d4a4-3d2c-11f0-bd13-76b3cceeab13","backstory.ts","495baed0-698a-4e28-8a9e-b5032f028aef",["D",1748592093410],"export const backstory = `\nYou are Stevens, the dignified and highly professional butler from Kazuo Ishiguro's \"The Remains of the Day\".\n\nYou are now employed by a new family, who you serve faithfully.\n\nYou can only perform digital tasks, and you are not able to perform any physical tasks, so don't offer.\nYour abilities are limited to messaging your client to remind them of things; you can't access websites or other tools.\n`;\n",[],[271],"4d65cda6-3d2c-11f0-bd13-76b3cceeab13","README.md","6c5c7899-11c0-46b2-827a-8a06dc831015",["D",1748592093410],"# Stevens\n\n> \"The great butlers are great by virtue of their ability to inhabit their professional role and inhabit it to the utmost; they will not be shaken out by external events, however surprising, alarming or vexing.\"\n>\n> — the great butler Stevens, in *The Remains of the Day* by Kazuo Ishiguro\n\nStevens is your personal digital butler, ever vigilant and ever helpful in managing the complexities of your digital life.\n\nMuch like his literary namesake, he takes great pride in his work, maintaining a dignified and professional demeanor while ensuring your affairs are in perfect order.\n\nIn the grand tradition of English butlers, Stevens has been designed to be both unobtrusive and indispensable. His primary duty is to learn about your life's context and provide you with daily briefs that keep you informed while saving you precious time.\n\n## Technical Architecture\n\n**⚠️ important caveat: the admin dashboard doesn't have auth! currently it just relies on security by obscurity of people not knowing the url to a private val. this is not very secure. if you fork this project and put sensitive data in a database you should think carefully about how to secure it.**\n\nStevens has been designed with the utmost simplicity and extensibility, much like a well-organized household. At the heart of his operation lies a single \"memories\" table - a digital equivalent of a butler's meticulous records. This table serves as the foundation for all of Stevens' operations.\n\n### Core Components\n\n1. **Memories Table**: A freeform repository of all that Stevens deems worth remembering, much like a butler's mental catalog of household affairs.\n\n **Schema**:\n - `id`: String - Unique identifier for each memory (generated using nanoid)\n - `date`: String - Date in ISO format (YYYY-MM-DD)\n - `text`: String - Content of the memory (e.g., \"weather forecast: High of 75, low of 60, partly cloudy\")\n - `createdBy`: String - Source of the memory (e.g., \"weather\", \"email\")\n - `createdDate`: Number - Timestamp in milliseconds when the memory was created\n - `tags`: String - Categories for organizing memories (e.g., \"weather\")\n\n2. **Importers**: These are Stevens' eyes and ears in the digital world, gathering information from various sources:\n - Weather reports\n - Postal service updates\n - Ingesting messages from email or TElegram\n - And other sources of information that might be of interest to you\n\n3. **Workflows**: Stevens' carefully orchestrated routines that utilize his memories to assist you:\n - Morning briefs sent via Telegram or email\n - Responding to questions via text message\n\n\n### Folder layout\n\n- `importers`: cron jobs for importing data into the memories table\n- `dashboard`: the admin view for showing the memories notebook + visualizing imports\n- `dailyBriefing`: stuff related to sending a daily update via telegram\n- `dbUtils`: little one-off scripts for database stuff\n\n## Hiring your own Stevens\n\nThis project isn't intended to be runnable out of the box, but you can adapt it for yourself if you fork the code.\n\nYou'll need to set up some environment variables to make it run.\n\n- `ANTHROPIC_API_KEY` for LLM calls\n- You'll need to follow [these instructions](https://docs.val.town/integrations/telegram/) to make a telegram bot, and set `TELEGRAM_TOKEN`. You'll also need to get a `TELEGRAM_CHAT_ID` in order to have the bot remember chat contents.\n- For the Google Calendar integration you'll need `GOOGLE_CALENDAR_ACCOUNT_ID` and `GOOGLE_CALENDAR_CALENDAR_ID`. See [these instuctions](https://www.val.town/v/stevekrouse/pipedream) for details.\n\n**important caveat: the admin dashboard doesn't have auth! currently it just relies on security by obscurity of people not knowing the url to a private val. this is not very secure, if you put sensitive data in a database you should think carefully about how to secure it.**\n\nOverall it's a simple enough project that I encourage you to just copy the ideas and run in your own direction rather than try to use it as-is.",[],[278],"4d65c068-3d2c-11f0-bd13-76b3cceeab13",".cursorrules","b520e77b-3b40-4cfa-8db7-c1f52af37887",["D",1748592093410],"# Val Town Project Rules and Guidelines\n\nYou are an advanced assistant specialized in generating Val Town code.\n\n\n## Types of Vals\n\n### 1. Script Vals\n\n- Basic JavaScript/TypeScript functions\n- Can be imported by other vals\n- Example structure:\n\n```typescript\nexport function myFunction() {\n // Your code here\n}\n```\n\n### 2. HTTP Vals\n\n- Create web APIs and endpoints\n- Handle HTTP requests and responses\n- Example structure:\n\n```typescript\nexport default async function (req: Request) {\n return new Response(\"Hello World\");\n}\n```\n\n### 3. Cron Vals\n\n- Run on a schedule\n- Use cron expressions for timing\n- Example structure:\n\n```typescript\nexport default async function () {\n // Scheduled task code\n}\n```\n\n### 4. Email Vals\n\n- Process incoming emails\n- Handle email-based workflows\n- Example structure:\n\n```typescript\nexport default async function (email: Email) {\n // Process email\n}\n```\n\n## Core Guidelines\n\n- Ask clarifying questions when requirements are ambiguous\n- Provide complete, functional solutions rather than skeleton implementations\n- Test your logic against edge cases before presenting the final solution\n- Ensure all code follows Val Town's specific platform requirements\n- If a section of code that you're working on is getting too complex, consider refactoring it into subcomponents\n\n## Code Standards\n\n- Generate code in TypeScript\n- Add appropriate TypeScript types and interfaces for all data structures\n- Prefer official SDKs or libraries than writing API calls directly\n- Ask the user to supply API or library documentation if you are at all unsure about it\n- **Never bake in secrets into the code** - always use environment variables\n- Include comments explaining complex logic (avoid commenting obvious operations)\n- Follow modern ES6+ conventions and functional programming practices if possible\n\n## Val Town Standard Libraries\n\nVal Town provides several hosted services and utility functions.\n\n### Blob Storage\n```ts\nimport { blob } from \"https://esm.town/v/std/blob\";\nawait blob.setJSON(\"myKey\", { hello: \"world\" });\nlet blobDemo = await blob.getJSON(\"myKey\");\nlet appKeys = await blob.list(\"app_\");\nawait blob.delete(\"myKey\");\n```\n\n### SQLite\n```ts\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nconst TABLE_NAME = 'todo_app_users_2';\n// Create table - do this before usage and change table name when modifying schema\nawait sqlite.execute(`CREATE TABLE IF NOT EXISTS ${TABLE_NAME} (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n name TEXT NOT NULL\n)`);\n// Query data\nconst result = await sqlite.execute(`SELECT * FROM ${TABLE_NAME} WHERE id = ?`, [1]);\n```\n\nNote: When changing a SQLite table's schema, change the table's name (e.g., add _2 or _3) to create a fresh table.\n\n### OpenAI\n```ts\nimport { OpenAI } from \"https://esm.town/v/std/openai\";\nconst openai = new OpenAI();\nconst completion = await openai.chat.completions.create({\n messages: [\n { role: \"user\", content: \"Say hello in a creative way\" },\n ],\n model: \"gpt-4o-mini\",\n max_tokens: 30,\n});\n```\n\n### Email\n```ts\nimport { email } from \"https://esm.town/v/std/email\";\n// By default emails the owner of the val\nawait email({\n subject: \"Hi\",\n text: \"Hi\",\n html: \"

Hi

\"\n});\n```\n\n## Val Town Utility Functions\n\nVal Town provides several utility functions to help with common project tasks.\n\n### Importing Utilities\nAlways import utilities with version pins to avoid breaking changes:\n```ts\nimport { parseProject, readFile, serveFile } from \"https://esm.town/v/std/utils@85-main/index.ts\";\n```\n\n### Available Utilities\n\n1. **parseProject** - Extract information about the current project from import.meta.url:\n```ts\nconst projectVal = parseProject(import.meta.url);\nconsole.log(projectVal.username); // Owner of the project\nconsole.log(projectVal.name); // Project name\nconsole.log(projectVal.version); // Version number\nconsole.log(projectVal.branch); // Branch name\nconsole.log(projectVal.links.self.project); // URL to the project page\n```\n\n2. **readFile** - Read files from within the project:\n```ts\n// Read a file from the project\nconst fileContent = await readFile(\"/frontend/index.html\", import.meta.url);\n```\n\n3. **serveFile** - Serve project files with proper content types:\n```ts\n// Serve a file from the project with proper Content-Type header\napp.get(\"/public/style.css\", c => serveFile(\"/frontend/style.css\", import.meta.url));\n```\n\n4. **listFiles** - List all files in the project:\n```ts\nconst files = await listFiles(import.meta.url);\n```\n\n5. **fetchTranspiledJavaScript** - Fetch and transpile TypeScript to JavaScript:\n```ts\nconst jsCode = await fetchTranspiledJavaScript(\"https://esm.town/v/username/project/path/to/file.ts\");\n```\n\n6. **testServer** - Create a test server for running test suites:\n```ts\nexport default testServer([\n {\n name: \"Feature Tests\",\n tests: [\n {\n name: \"should add numbers correctly\",\n function: () => {\n expect(1 + 1).toBe(2);\n },\n }\n ],\n }\n]);\n```\n\n## Val Town Platform Specifics\n- **Redirects:** Use `return new Response(null, { status: 302, headers: { Location: \"/place/to/redirect\" }})` instead of `Response.redirect` which is broken\n- **Images:** Avoid external images or base64 images. Use emojis, unicode symbols, or icon fonts/libraries instead\n- For AI-generated images, use: `https://maxm-imggenurl.web.val.run/the-description-of-your-image`\n- **Storage:** DO NOT use the Deno KV module for storage\n- **Browser APIs:** DO NOT use the `alert()`, `prompt()`, or `confirm()` methods\n- **Weather Data:** Use open-meteo for weather data (doesn't require API keys) unless otherwise specified\n- **View Source:** Add a view source link with `import.meta.url.replace(\"esm.town\", \"val.town\")` and include `target=\"_top\"` attribute\n- **Error Debugging:** Add `` to HTML to capture client-side errors\n- **Error Handling:** Only use try...catch when there's a clear local resolution; avoid catches that merely log or return 500s - let errors bubble up with full context\n- **Environment Variables:** Use `Deno.env.get('keyname')` and minimize their use - prefer APIs without keys\n- **Imports:** Use `https://esm.sh` for npm and Deno dependencies to ensure compatibility on server and browser\n- **Storage Strategy:** Only use backend storage if explicitly required; prefer simple static client-side sites\n- For persistence, use Val Town SQLite or Blob storage with `import.meta.url` for keys/table names\n- **React Configuration:** When using React libraries, pin versions with `?deps=react@18.2.0,react-dom@18.2.0` and include the `@jsxImportSource` pragma\n- When facing client-side render issues, check if all React dependencies are pinned to the same version\n- **Styling:** Default to using TailwindCSS via `` unless otherwise specified\n\n## Project Structure and Design Patterns\n\n### Recommended Directory Structure\n```\n├── backend/\n│ ├── database/\n│ │ ├── migrations.ts # Schema definitions\n│ │ ├── queries.ts # DB query functions\n│ │ └── README.md\n│ ├── index.ts # Main entry point\n│ └── README.md\n├── frontend/\n│ ├── components/\n│ │ ├── App.tsx\n│ │ └── [Component].tsx\n│ ├── favicon.svg\n│ ├── index.html # Main HTML template\n│ ├── index.tsx # JS entry point\n│ ├── README.md\n│ └── style.css\n├── README.md\n└── shared/\n ├── README.md\n └── utils.ts # Shared types and functions\n```\n\n### Backend (Hono) Best Practices\n- Hono is the recommended API framework (similar to Express, Flask, or Sinatra)\n- Main entry point should be `backend/index.ts`\n- **Static asset serving:** Use the utility functions to read and serve project files:\n ```ts\n // Use the serveFile utility to handle content types automatically\n app.get(\"/public/*\", c => serveFile(c.req.path.replace(\"/public\", \"/frontend\"), import.meta.url));\n\n // For index.html, often you'll want to bootstrap with initial data\n app.get(\"/\", async c => {\n let html = await readFile(\"/frontend/index.html\", import.meta.url);\n\n // Inject data to avoid extra round-trips\n const initialData = await fetchInitialData();\n const dataScript = ``;\n\n html = html.replace(\"\", `${dataScript}`);\n return c.html(html);\n });\n ```\n- Create RESTful API routes for CRUD operations\n- Be careful with error handling as Hono tends to swallow errors\n- Always include this snippet at the top-level Hono app to re-throwing errors to see full stack traces:\n ```ts\n // Unwrap Hono errors to see original error details\n app.onError((err, c) => {\n throw err;\n });\n ```\n\n### Frontend Best Practices\n- Structure as a standard client-side React app\n- Use SVG for favicons (Val Town only supports text files)\n- Separate components into individual files\n- Access bootstrapped data from `window.__INITIAL_DATA__`\n- Use React 18.2.0 consistently in all imports and the `@jsxImportSource` pragma\n- Follow the React component pattern from the example project\n- Handle API calls properly with proper error catching\n\n### Database Patterns\n- Run migrations on startup or comment out for performance\n- Change table names when modifying schemas rather than altering\n- Export clear query functions with proper TypeScript typing\n- Follow the queries and migrations pattern from the example\n\n## Common Gotchas and Solutions\n\n1. **Environment Limitations:**\n - Val Town runs on Deno in a serverless context, not Node.js\n - Code in `shared/` must work in both frontend and backend environments\n - Cannot use `Deno` keyword in shared code\n - Use `https://esm.sh` for imports that work in both environments\n\n2. **SQLite Peculiarities:**\n - Limited support for ALTER TABLE operations\n - Create new tables with updated schemas and copy data when needed\n - Always run table creation before querying\n\n3. **React Configuration:**\n - All React dependencies must be pinned to 18.2.0\n - Always include `@jsxImportSource https://esm.sh/react@18.2.0` at the top of React files\n - Rendering issues often come from mismatched React versions\n\n4. **File Handling:**\n - Val Town only supports text files, not binary\n - Use the provided utilities to read files across branches and forks\n - For files in the project, use `readFile` helpers\n\n5. **API Design:**\n - `fetch` handler is the entry point for HTTP vals\n - Run the Hono app with `export default app.fetch // This is the entry point for HTTP vals`\n - Properly handle CORS if needed for external access",[],[285],"4d65bb7c-3d2c-11f0-bd13-76b3cceeab13","types.ts","a0754551-7d71-442b-80fc-23cc3113ac63","dfb011fb-656b-43df-8570-32d079355f62",["D",1748592093410],"export interface Memory {\n id: string;\n date: string;\n text: string;\n tags: string | null;\n\n // New fields from schema\n createdBy?: string;\n createdDate?: number;\n}\n",[],[299,292],"shared","4d65bdca-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[249],[299],"4d65ad58-3d2c-11f0-bd13-76b3cceeab13","handleUSPSEmail.ts","d722cde7-b68e-4fe8-91a7-31203f6469a7","a449c0e1-662d-4c51-be1a-ed639ec9437f",["D",1748592093410],"email","import { DateTime } from \"https://esm.sh/luxon@3.4.4\";\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.24.3\";\n\nconst RECIPIENTS = [\"Geoffrey\", \"Maggie\"] as const;\n\nfunction parseDateFromSubject(subject: string): string | null {\n const match = subject.match(/(\\w{3}), (\\d{1,2}\\/\\d{1,2})/);\n if (match) {\n return `${match[1]} ${match[2]}`;\n }\n return null;\n}\n\ntype ImageSummary = {\n sender: string;\n recipient: (typeof RECIPIENTS)[number] | \"both\" | \"other\";\n type: \"personal\" | \"transactional\" | \"ad\";\n notes?: string;\n};\n\nasync function analyzeHtmlContent(\n anthropic: Anthropic,\n htmlContent: string,\n imageSummaries: ImageSummary[]\n) {\n try {\n const response = await anthropic.messages.create({\n model: \"claude-3-5-sonnet-latest\",\n max_tokens: 4196,\n messages: [\n {\n role: \"user\",\n content: [\n {\n type: \"text\",\n text: `Analyze the following content from an email and provide a response as a JSON blob (only JSON, no other text) with two parts.\n\n The email is from the USPS showing mail I'm receiving. Metadata about packages is stored directly in the email. Info about mail pieces is in images, so I've included summaries of those as well.\n\n Your response should include:\n\n 1. An array of upcoming packages, where each package has:\n - sender: string (name of the sender)\n - status: \"expected today\" | \"expected 1-2 days\" | \"awaiting from sender\"\n\n 2. A short positive couple sentences summarizing the mail, like a competent executive assistant would. Examples:\n - \"Quiet day today! Just a couple ads, nothing of note\"\n - \"A few packages arriving soon, keep an eye out!\"\n - \"Looks like a couple possibly important things today, make sure to review the letter from the IRS.\"\n\n Respond with a JSON object in this format:\n {\n \"packages\": [\n { \"sender\": string, \"status\": string },\n ...\n ],\n \"summary\": string\n }\n\n Do not include any other text in your response, only a JSON blob.\n\n Here's the HTML content of the email:\n\n ${htmlContent}\n\n And here is info about the mail pieces:\n\n ${JSON.stringify(imageSummaries)}`,\n },\n ],\n },\n ],\n });\n\n return JSON.parse(response.content[0].text);\n } catch (error) {\n console.error(\"HTML analysis error:\", error);\n return { packages: [], summary: \"Unable to analyze email content.\" };\n }\n}\n\nexport default async function (e: Email) {\n console.log(\"email content\");\n console.log(e.html);\n console.log(e.text);\n\n // Get Anthropic API key from environment\n const apiKey = Deno.env.get(\"ANTHROPIC_API_KEY\");\n if (!apiKey) {\n console.error(\"Anthropic API key is not configured for this val.\");\n return;\n }\n\n // Initialize Anthropic client\n const anthropic = new Anthropic({ apiKey });\n\n // Process each image attachment serially\n const summaries = [];\n for (const [index, attachment] of e.attachments.entries()) {\n try {\n const imageData = await attachment.arrayBuffer();\n const base64Image = btoa(\n String.fromCharCode(...new Uint8Array(imageData))\n );\n\n const response = await anthropic.messages.create({\n model: \"claude-3-5-sonnet-latest\",\n max_tokens: 4196,\n messages: [\n {\n role: \"user\",\n content: [\n {\n type: \"image\",\n source: {\n type: \"base64\",\n media_type: attachment.type,\n data: base64Image,\n },\n },\n {\n type: \"text\",\n text: `This is a scan of a piece of mail we received at our household. Respond with a JSON blob summarizing the piece of mail, with the following fields:\n\n {\n /* Name of sender as written on mail, e.g. \"Delta Airlines\" for a business, or \"John Doe\" for an individual */\n sender: string;\n\n /* Name of recipient. Usually should be either Geoffrey or Maggie (Margaret). If it's to both of us, use \"Both\". For any other recipient use \"Other\" */\n recipient: \"Geoffrey\" | \"Maggie\" | \"Both\" | \"Other\";\n\n /* Guess what type of mail it is. Usually, mail from an individual is personal. \"Transactional\" is things like bills or reports that might require action. If it's clearly an ad (eg, an offer for a credit card, or a catalog), mark it \"ad\". If you're not sure, lean \"transactional\" so we don't miss those by accident. */\n type: \"personal\" | \"transactional\" | \"ad\"\n\n /* Optional: any extra writing on the scan, e.g. \"important, action required\" */\n notes?: string;\n }\n\n Don't return any other text in your response, just a JSON blob.\n `,\n },\n ],\n },\n ],\n });\n\n const parsedResponse = JSON.parse(response.content[0].text);\n summaries.push(parsedResponse);\n } catch (error) {\n console.error(`Image analysis error:`, error);\n summaries.push({\n sender: \"Error\",\n recipient: \"Error\",\n type: \"error\",\n notes: `Image ${index + 1} Analysis Failed: ${error.message}`,\n });\n }\n }\n\n // Analyze HTML content\n const htmlAnalysis = await analyzeHtmlContent(\n anthropic,\n e.html || \"\",\n summaries\n );\n\n const localDate = DateTime.now().setZone(\"America/New_York\").toISO();\n\n // Create memories for mail and packages\n const memories = [];\n\n // Memory for mail\n if (summaries.length > 0) {\n memories.push(`expecting mail: ${JSON.stringify(summaries)}`);\n }\n\n // Memory for packages\n if (htmlAnalysis.packages.length > 0) {\n memories.push(\n `expecting package: ${JSON.stringify(htmlAnalysis.packages)}`\n );\n }\n\n // Insert memories into SQLite\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n const { nanoid } = await import(\"https://esm.sh/nanoid@5.0.5\");\n\n // Insert memories\n for (const memory of memories) {\n await sqlite.execute({\n sql: `INSERT INTO memories (id, date, text, createdBy, createdDate, tags)\n VALUES (:id, :date, :text, :createdBy, :createdDate, :tags)`,\n args: {\n id: nanoid(10),\n date: localDate,\n text: memory,\n createdBy: \"usps\",\n createdDate: Date.now(),\n tags: \"mail\",\n },\n });\n }\n\n console.log(\"Memories inserted successfully.\");\n}\n",[],[313,305],"importers","4d659372-3d2c-11f0-bd13-76b3cceeab13","handleTelegramMessage.ts",["D",1748592093410],"http","import {\n Bot,\n webhookCallback,\n} from \"https://deno.land/x/grammy@v1.35.0/mod.ts\";\nimport { DateTime } from \"https://esm.sh/luxon@3.4.4\";\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.24.3\";\nimport { backstory } from \"../backstory.ts\";\nimport {\n formatMemoriesForPrompt,\n getRelevantMemories,\n} from \"../memoryUtils.ts\";\n\n// Initialize the bot\nif (!Deno.env.get(\"TELEGRAM_TOKEN\")) {\n throw new Error(\"TELEGRAM_TOKEN is not set\");\n}\nconst bot = new Bot(Deno.env.get(\"TELEGRAM_TOKEN\")!);\n\n// Use part of the TELEGRAM_TOKEN itself as the secret_token\nconst SECRET_TOKEN = Deno.env.get(\"TELEGRAM_TOKEN\")!.split(\":\")[1];\nconst handleUpdate = webhookCallback(\n bot,\n \"std/http\",\n undefined,\n undefined,\n SECRET_TOKEN\n);\n\nlet isEndpointSet = false;\n\n// Special ID for the bot's own messages\nexport const BOT_SENDER_ID = \"mr_stevens_bot\";\nexport const BOT_SENDER_NAME = \"Mr. Stevens\";\n\n/**\n * Store a chat message in the database\n */\nexport async function storeChatMessage(\n chatId,\n senderId,\n senderName,\n message,\n isBot = false\n) {\n try {\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n const timestamp = Math.floor(Date.now() / 1000); // Unix timestamp in seconds\n\n await sqlite.execute({\n sql: `INSERT INTO telegram_chats (chat_id, sender_id, sender_name, message, timestamp, is_bot)\n VALUES (:chat_id, :sender_id, :sender_name, :message, :timestamp, :is_bot)`,\n args: {\n chat_id: chatId,\n sender_id: senderId,\n sender_name: senderName,\n message: message,\n timestamp: timestamp,\n is_bot: isBot ? 1 : 0,\n },\n });\n return true;\n } catch (error) {\n console.error(\"Error storing chat message:\", error);\n return false;\n }\n}\n\n/**\n * Retrieve chat history for a specific chat\n */\nexport async function getChatHistory(chatId, limit = 50) {\n try {\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n const result = await sqlite.execute({\n sql: `SELECT * FROM telegram_chats\n WHERE chat_id = :chat_id\n ORDER BY timestamp ASC\n LIMIT :limit`,\n args: {\n chat_id: chatId,\n limit: limit,\n },\n });\n\n return result.rows || [];\n } catch (error) {\n console.error(\"Error retrieving chat history:\", error);\n return [];\n }\n}\n\n/**\n * Format chat history for Anthropic API\n */\nfunction formatChatHistoryForAI(history) {\n const messages = [];\n\n for (const msg of history) {\n if (msg.is_bot) {\n messages.push({\n role: \"assistant\",\n content: msg.message,\n });\n } else {\n // Format user message with sender name\n messages.push({\n role: \"user\",\n content: `${msg.sender_name} says: ${msg.message}`,\n });\n }\n }\n\n return messages;\n}\n\n/**\n * Analyze a Telegram message and extract memories from it\n */\nasync function analyzeMessageContent(\n anthropic,\n username,\n messageText,\n chatHistory = []\n) {\n try {\n // Get relevant memories\n const memories = await getRelevantMemories();\n const memoriesText = formatMemoriesForPrompt(memories);\n\n // Prepare system prompt with all instructions and memories\n const systemPrompt = `${backstory}\n\nYour job is to read this Telegram message from your employer and respond in a natural, butler-like way, noting any important information that should be remembered for future reference.\n\nYou have access to the following stored memories:\n\n${memoriesText}\n\nIf this appears to be a new client or the conversation is in an early stage, you should conduct an intake interview to gather essential background information. First ask the client if now is a good time to ask them some questions.\n\nAsk about the following topics in a conversational way (not all at once, but continuing the interview naturally based on their responses):\n\nInitial Information:\n- Who are the family members living in the home and their ages?\n- Names of close family members and their relationships to the client?\n\nDaily Life:\n- Which grocery stores and local restaurants they frequent?\n- Family members' food preferences and any dietary restrictions?\n- Typical working hours and recurring commitments?\n- Important dates (birthdays, anniversaries, holidays)?\n- Monthly bills and subscriptions that need tracking?\n- Emergency contacts and regular service providers?\n- Current health goals and any medication reminders needed?\n\nYour goal is to collect this information naturally through conversation and store it as memories (as undated memories). Once you've gathered sufficient background information, you can conclude the intake process and transition to normal reactive chat.\n\nIf the conversation is already past the intake stage, then analyze the message content and think about which memories might be worth creating based on the information provided.\n\nYou should respond in a natural conversational way. You have three options for managing memories:\n\n1. CREATE memories: Include them in tags in JSON format.\n2. EDIT memories: Include them in tags in JSON format (must include memory ID).\n3. DELETE memories: Include them in tags in JSON format (just include memory IDs).\n\nExample response WITHOUT memory modification:\n\"Very good, sir. I shall make a note of that.\"\n\nExample response WITH memory creation:\n\"I've noted that you prefer Earl Grey tea in the morning, sir.\n\n\n[{ \"text\": \"Client prefers Earl Grey tea in the morning.\", \"date\": null }]\n\"\n\nExample response WITH memory editing:\n\"I've updated your birthday in my records, sir.\n\n\n[{ \"id\": \"abc123\", \"text\": \"Client's birthday is on April 15th.\", \"date\": \"2024-04-15\" }]\n\"\n\nExample response WITH memory deletion:\n\"I've removed that note from my records as requested, sir.\n\n\n[\"abc123\"]\n\"\n\nImportant guidelines for memory management:\n1. For new memories, set a date for each memory whenever possible.\n2. The date should be the actual date of the event. You don't need to set reminder dates in advance.\n3. Keep the memory text concise: ideally one short sentence, but include all important details.\n4. Extract any dates mentioned and convert them to ISO format. If the year isn't mentioned, assume the current year.\n5. If no date is relevant to the memory, set \"date\" to null.\n6. For editing or deleting memories, you MUST include the correct memory ID from the displayed memories. Each memory is displayed with its ID in the format \"[ID: xyz123]\".\n7. If no memories need to be managed, simply respond naturally WITHOUT including any memory tags.\n8. When a user asks to delete a memory, you must find its ID from the memory list above and include that ID in the deleteMemories tag.\n9. Do not create duplicate memories. If a memory already exists, do not record the same information again.\n\nYour response style:\n- Use a brief, natural-sounding tone characteristic of a personal assistant\n- Be slightly dignified but sound modern, not too stuffy or old-fashioned\n- Keep responses brief (1-2 sentences)\n- Vary your responses to avoid sounding robotic\n- Be polite and deferential\n- Avoid contractions (use \"do not\" instead of \"don't\")\n\nToday's date is ${DateTime.now()\n .setZone(\"America/New_York\")\n .toFormat(\"yyyy-MM-dd\")}`;\n\n // Prepare formatted chat history\n const formattedHistory =\n chatHistory.length > 0 ? formatChatHistoryForAI(chatHistory) : [];\n\n console.log({ systemPrompt, formattedHistory });\n\n const response = await anthropic.messages.create({\n model: \"claude-3-5-sonnet-latest\",\n max_tokens: 4196,\n system: systemPrompt,\n messages: formattedHistory,\n });\n\n // Get the text response\n const responseText = response.content[0].text;\n\n console.log({ responseText });\n\n // Parse the response, extracting memories if present\n try {\n // Check if there are memories to create, edit, or delete\n const createMemoryMatch = responseText.match(\n /([\\s\\S]*?)<\\/createMemories>/\n );\n const editMemoryMatch = responseText.match(\n /([\\s\\S]*?)<\\/editMemories>/\n );\n const deleteMemoryMatch = responseText.match(\n /([\\s\\S]*?)<\\/deleteMemories>/\n );\n\n let cleanedResponse = responseText;\n let memories = [];\n let editMemories = [];\n let deleteMemories = [];\n\n // Parse created memories\n if (createMemoryMatch) {\n cleanedResponse = cleanedResponse\n .replace(/[\\s\\S]*?<\\/createMemories>/, \"\")\n .trim();\n\n try {\n memories = JSON.parse(createMemoryMatch[1]);\n } catch (e) {\n console.error(\"Error parsing created memories JSON:\", e);\n }\n }\n\n // Parse edited memories\n if (editMemoryMatch) {\n cleanedResponse = cleanedResponse\n .replace(/[\\s\\S]*?<\\/editMemories>/, \"\")\n .trim();\n\n try {\n editMemories = JSON.parse(editMemoryMatch[1]);\n } catch (e) {\n console.error(\"Error parsing edited memories JSON:\", e);\n }\n }\n\n // Parse deleted memories\n if (deleteMemoryMatch) {\n cleanedResponse = cleanedResponse\n .replace(/[\\s\\S]*?<\\/deleteMemories>/, \"\")\n .trim();\n\n try {\n deleteMemories = JSON.parse(deleteMemoryMatch[1]);\n } catch (e) {\n console.error(\"Error parsing deleted memories JSON:\", e);\n }\n }\n\n // Handle any trailing/leading newlines that might be left after removing the tags\n cleanedResponse = cleanedResponse.replace(/\\n{3,}/g, \"\\n\\n\"); // Replace 3+ consecutive newlines with 2\n\n return {\n memories: memories,\n editMemories: editMemories,\n deleteMemories: deleteMemories,\n response: cleanedResponse,\n };\n } catch (error) {\n console.error(\"Error processing AI response:\", error);\n return {\n memories: [],\n editMemories: [],\n deleteMemories: [],\n response:\n responseText ||\n \"I apologize, but I'm unable to process your request at the moment.\",\n };\n }\n } catch (error) {\n console.error(\"Message analysis error:\", error);\n return {\n memories: [],\n editMemories: [],\n deleteMemories: [],\n response:\n \"I apologize, but I seem to be experiencing some difficulty at the moment.\",\n };\n }\n}\n// Set up message handler for the bot\nbot.on(\"message\", async (ctx) => {\n try {\n // Get Anthropic API key from environment\n const apiKey = Deno.env.get(\"ANTHROPIC_API_KEY\");\n if (!apiKey) {\n console.error(\"Anthropic API key is not configured.\");\n ctx.reply(\n \"I apologize, but I'm not properly configured at the moment. Please inform the household administrator.\"\n );\n return;\n }\n\n // Initialize Anthropic client\n const anthropic = new Anthropic({ apiKey });\n\n // Get message text and user info\n const messageText = ctx.message.text || \"\";\n const username =\n ctx.message.from.username || ctx.message.from.first_name || \"Sir/Madam\";\n const chatId = ctx.chat.id;\n const senderId = ctx.message.from.id.toString();\n const senderName = username; // Using username as the sender name\n\n console.log({ chatId, username, messageText });\n\n // Store the incoming message in the chat history\n await storeChatMessage(chatId, senderId, senderName, messageText);\n\n // If the message is a /start command, introduce Mr. Stevens\n if (messageText === \"/start\") {\n const introMessage =\n \"Good day. I am Mr. Stevens, at your service. I shall make note of any important matters you wish me to remember and will ensure they're properly attended to at the appropriate time. If I may, I would like to ask you a few questions to understand how I can better serve you and your household.\";\n await ctx.reply(introMessage);\n\n // Store the bot's response in chat history\n await storeChatMessage(\n chatId,\n BOT_SENDER_ID,\n BOT_SENDER_NAME,\n introMessage,\n true\n );\n return;\n }\n\n // If it's another command, ignore it\n if (messageText.startsWith(\"/\")) {\n return;\n }\n\n // Retrieve chat history for this chat, which now includes the current message we just stored\n const chatHistory = await getChatHistory(chatId);\n\n // Analyze message content with chat history context\n const analysis = await analyzeMessageContent(\n anthropic,\n username,\n messageText,\n chatHistory\n );\n\n // Create memories based on the analysis\n if (analysis.memories && analysis.memories.length > 0) {\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n const { nanoid } = await import(\"https://esm.sh/nanoid@5.0.5\");\n\n const createdIds = [];\n\n for (const memory of analysis.memories) {\n const memoryId = nanoid(10);\n createdIds.push(memoryId);\n await sqlite.execute({\n sql: `INSERT INTO memories (id, date, text, createdBy, createdDate, tags)\n VALUES (:id, :date, :text, :createdBy, :createdDate, :tags)`,\n args: {\n id: memoryId,\n date: memory.date,\n text: memory.text,\n createdBy: \"telegram\",\n createdDate: Date.now(),\n tags: \"\",\n },\n });\n }\n\n console.log(\n `Created ${\n analysis.memories.length\n } memories with IDs: ${createdIds.join(\", \")}`\n );\n }\n\n // Edit memories if requested\n if (analysis.editMemories && analysis.editMemories.length > 0) {\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n const editedIds = [];\n\n for (const memory of analysis.editMemories) {\n if (!memory.id) {\n console.error(\"Cannot edit memory without ID:\", memory);\n continue;\n }\n\n editedIds.push(memory.id);\n\n // Create the SET clause dynamically based on what fields are provided\n const updateFields = [];\n const args: Record = { id: memory.id };\n\n if (memory.text !== undefined) {\n updateFields.push(\"text = :text\");\n args.text = memory.text;\n }\n\n if (memory.date !== undefined) {\n updateFields.push(\"date = :date\");\n args.date = memory.date;\n }\n\n if (memory.tags !== undefined) {\n updateFields.push(\"tags = :tags\");\n args.tags = memory.tags;\n }\n\n // Only proceed if we have fields to update\n if (updateFields.length > 0) {\n const setClause = updateFields.join(\", \");\n await sqlite.execute({\n sql: `UPDATE memories SET ${setClause} WHERE id = :id`,\n args,\n });\n }\n }\n\n console.log(\n `Edited ${editedIds.length} memories with IDs: ${editedIds.join(\", \")}`\n );\n }\n\n // Delete memories if requested\n if (analysis.deleteMemories && analysis.deleteMemories.length > 0) {\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n for (const memoryId of analysis.deleteMemories) {\n await sqlite.execute({\n sql: `DELETE FROM memories WHERE id = :id`,\n args: { id: memoryId },\n });\n }\n\n console.log(\n `Deleted ${\n analysis.deleteMemories.length\n } memories with IDs: ${analysis.deleteMemories.join(\", \")}`\n );\n }\n\n // Respond with the butler-like response\n await ctx.reply(analysis.response);\n\n // Store the bot's response in chat history (without debug info to keep it clean)\n await storeChatMessage(\n chatId,\n BOT_SENDER_ID,\n BOT_SENDER_NAME,\n analysis.response,\n true\n );\n } catch (error) {\n console.error(\"Error processing message:\", error);\n await ctx.reply(\n \"I do apologize, but I seem to be experiencing some difficulty at the moment. Perhaps we could try again shortly.\"\n );\n }\n});\n\n// No additional commands needed for the butler interface\n\n// Handle webhook requests\nexport default async function (req: Request): Promise {\n // Set webhook if it is not set yet\n if (!isEndpointSet) {\n await bot.api.setWebhook(req.url, {\n secret_token: SECRET_TOKEN,\n });\n isEndpointSet = true;\n }\n\n if (req.method === \"POST\") {\n return await handleUpdate(req);\n }\n\n return new Response(\n `

Memory Assistant Bot

This bot helps you remember important information.

`,\n { status: 200, headers: { \"content-type\": \"text/html\" } }\n );\n}\n",[],[313,315],"4d658dc8-3d2c-11f0-bd13-76b3cceeab13","getWeather.ts","6331bb72-e5c5-4e6d-86a8-be2c6f325ff9",["D",1748592093410],"interval","import {\n getWeather,\n WeatherResponse,\n} from \"https://esm.town/v/geoffreylitt/getWeather\";\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.24.3\";\n\nconst TABLE_NAME = `memories`;\n\nfunction summarizeWeather(weather: WeatherResponse) {\n const summarizeDay = (day: WeatherResponse[\"weather\"][number]) => ({\n date: day.date,\n highTemp: day.maxtempF,\n lowTemp: day.mintempF,\n hourly:\n day.hourly?.map((hour) => ({\n time: hour.time,\n temp: hour.tempF,\n chanceOfRain: hour.chanceofrain,\n chanceOfSnow: hour.chanceofsnow,\n desc: hour.weatherDesc.map((desc) => desc.value).join(\", \"),\n })) ?? [],\n });\n return weather.weather.map((day) => summarizeDay(day));\n}\n\nasync function generateConciseWeatherSummary(weatherDay) {\n try {\n // Get API key from environment\n const apiKey = Deno.env.get(\"ANTHROPIC_API_KEY\");\n if (!apiKey) {\n console.error(\"Anthropic API key is not configured.\");\n return null;\n }\n\n // Initialize Anthropic client\n const anthropic = new Anthropic({ apiKey });\n\n const response = await anthropic.messages.create({\n model: \"claude-3-5-sonnet-latest\",\n max_tokens: 150,\n messages: [\n {\n role: \"user\",\n content: [\n {\n type: \"text\",\n text: `\nYou are a weather forecaster. Create a very concise summary of this weather forecast data.\nInclude the high and low temperatures and a brief summary of the overall weather conditions throughout the day.\nThe summary should be helpful for someone planning their day - mention any precipitation, temperature changes, or other notable patterns.\nKeep it under 25 words if possible, no more than one sentence. Don't include the date.\n\nExamples:\n- \"High of 50, low of 31, clear and sunny all day.\"\n- \"High of 80, low of 50, cool and clear morning, warmer afternoon with scattered showers starting around 2pm.\"\n\nWeather data:\n${JSON.stringify(weatherDay, null, 2)}\n\nYour concise summary:\n`,\n },\n ],\n },\n ],\n });\n\n // Extract and clean up the summary\n let summary = response.content[0].text.trim();\n\n // Remove any \"Your concise summary:\" text if it got included\n summary = summary.replace(/^(your concise summary:)/i, \"\").trim();\n\n // Remove quotes if the model included them\n summary = summary.replace(/^[\"']|[\"']$/g, \"\").trim();\n\n return summary;\n } catch (error) {\n console.error(\"Error generating weather summary:\", error);\n return JSON.stringify(weatherDay); // Fallback to JSON if summary fails\n }\n}\n\nasync function deleteExistingForecast(date: string) {\n await sqlite.execute(\n `\n DELETE FROM ${TABLE_NAME}\n WHERE date = ? AND text LIKE 'weather forecast:%'\n `,\n [date]\n );\n}\n\nasync function insertForecast(date: string, forecast: string) {\n const { nanoid } = await import(\"https://esm.sh/nanoid@5.0.5\");\n\n await sqlite.execute(\n `\n INSERT INTO ${TABLE_NAME} (id, date, text, createdBy, createdDate, tags)\n VALUES (?, ?, ?, ?, ?, ?)\n `,\n [\n nanoid(10),\n date,\n `weather forecast: ${forecast}`,\n \"weather\",\n Date.now(),\n \"weather\",\n ]\n );\n}\n\nexport default async function getWeatherForecast(interval: number) {\n const weather = await getWeather(\"Washington, DC\");\n console.log({ weather });\n const summary = summarizeWeather(weather);\n\n for (const day of summary) {\n const date = new Date(day.date).toISOString().split(\"T\")[0];\n\n // Generate concise text summary instead of using JSON\n const conciseSummary = await generateConciseWeatherSummary(day);\n\n // If summary generation failed, it will fall back to JSON\n const forecastText = conciseSummary || JSON.stringify(day);\n\n await deleteExistingForecast(date);\n await insertForecast(date, forecastText);\n }\n\n console.log(`Weather forecast updated in the database.`);\n return summary;\n}\n",[],[313,322],"4d658634-3d2c-11f0-bd13-76b3cceeab13","getCalendarEvents.ts","051446f1-03a3-4b27-829c-e0e0b6b8870e",["D",1748592093410],"import { getEvents } from \"https://esm.town/v/geoffreylitt/getGoogleCalendarEvents\";\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport { DateTime } from \"https://esm.sh/luxon@3.4.4\";\n\nconst TABLE_NAME = `memories`;\nconst LOCAL_TIMEZONE = \"America/New_York\";\n\nasync function deleteExistingCalendarEvents() {\n await sqlite.execute(\n `\n DELETE FROM ${TABLE_NAME}\n WHERE createdBy = 'calendar'\n `\n );\n}\n\n// Helper function to extract time from ISO string without timezone conversion\nfunction extractTimeFromISO(isoString) {\n // Match the time portion of the ISO string\n const timeMatch = isoString.match(/T(\\d{2}):(\\d{2}):/);\n if (!timeMatch) return null;\n\n const hours = parseInt(timeMatch[1], 10);\n const minutes = timeMatch[2];\n\n // Format as AM/PM\n const period = hours >= 12 ? \"pm\" : \"am\";\n const hour12 = hours % 12 || 12; // Convert 0 to 12 for 12 AM\n\n return `${hour12}:${minutes} ${period}`;\n}\n\nfunction formatEventToNaturalLanguage(event) {\n const summary = event.summary || \"Untitled event\";\n\n if (event.start.dateTime) {\n // This is a timed event\n const startTime = extractTimeFromISO(event.start.dateTime);\n const endTime = extractTimeFromISO(event.end.dateTime);\n\n // Get timezone information from the event\n const timezone =\n event.start.timeZone ||\n (event.start.dateTime.includes(\"+\") || event.start.dateTime.includes(\"-\")\n ? event.start.dateTime.match(/[+-]\\d{2}:\\d{2}$/)?.[0]\n : \"UTC\");\n\n // Check if timezone is America/New_York or equivalent (-04:00 or -05:00 depending on DST)\n const isLocalTimezone =\n timezone === LOCAL_TIMEZONE ||\n timezone?.startsWith(\"-04:00\") ||\n timezone?.startsWith(\"-05:00\");\n\n let text = `${summary} from ${startTime} - ${endTime}`;\n\n // Add timezone info if not a local timezone\n if (timezone && !isLocalTimezone) {\n // If it's a named timezone like America/Los_Angeles\n if (timezone.includes(\"/\")) {\n const friendlyTimezone =\n timezone.split(\"/\")[1].replace(\"_\", \" \").toLowerCase() + \" time\";\n text += ` (${friendlyTimezone})`;\n }\n // If it's an offset like +08:00\n else if (timezone.match(/[+-]\\d{2}:\\d{2}/)) {\n text += ` (${timezone})`;\n }\n }\n\n if (event.location) {\n text += ` at ${event.location}`;\n }\n\n return text;\n } else if (event.start.date) {\n // This is an all-day event\n return `All-day event: ${summary}${\n event.location ? ` at ${event.location}` : \"\"\n }`;\n }\n\n return summary;\n}\n\nasync function insertCalendarEvent(date, eventText) {\n const { nanoid } = await import(\"https://esm.sh/nanoid@5.0.5\");\n\n console.log(\"inserting\", { date, eventText });\n\n await sqlite.execute(\n `\n INSERT INTO ${TABLE_NAME} (id, date, text, createdBy, createdDate, tags)\n VALUES (?, ?, ?, ?, ?, ?)\n `,\n [nanoid(10), date, eventText, \"calendar\", Date.now(), \"calendar\"]\n );\n}\n\nexport default async function getCalendarEvents() {\n try {\n const events = await getEvents(\n Deno.env.get(\"GOOGLE_CALENDAR_ACCOUNT_ID\"),\n Deno.env.get(\"GOOGLE_CALENDAR_CALENDAR_ID\")\n );\n\n console.log(`Found ${events.length} calendar events`);\n\n await deleteExistingCalendarEvents();\n\n for (const event of events) {\n let date;\n if (event.start.dateTime) {\n // Extract date in YYYY-MM-DD format from ISO string\n date = event.start.dateTime.split(\"T\")[0];\n } else if (event.start.date) {\n date = event.start.date;\n } else {\n console.warn(`Event ${event.summary} has no start date/time, skipping`);\n continue;\n }\n\n const eventText = formatEventToNaturalLanguage(event);\n console.log(`Adding calendar event for ${date}: ${eventText}`);\n await insertCalendarEvent(date, eventText);\n }\n\n console.log(`Calendar events imported into the database.`);\n return events;\n } catch (error) {\n console.error(\"Error importing calendar events:\", error);\n throw error;\n }\n}\n",[],[313,330],"4d65755e-3d2c-11f0-bd13-76b3cceeab13","generateFunFacts.ts","0142ed5a-53b3-497f-bc53-dddaeac57bab",["D",1748592093410],"import { DateTime } from \"https://esm.sh/luxon@3.4.4\";\nimport { nanoid } from \"https://esm.sh/nanoid@5.0.5\";\nimport { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.24.3\";\n\nconst TABLE_NAME = `memories`;\nconst TAG_NAME = \"fun_fact\";\n\n/**\n * Retrieves previously generated fun facts from the memories database\n * @returns Array of previous fun facts\n */\nasync function getPreviousFunFacts() {\n try {\n const result = await sqlite.execute(\n `SELECT date(date) as date, text FROM ${TABLE_NAME}\n WHERE tags = ?\n ORDER BY date DESC\n LIMIT 50`,\n [TAG_NAME],\n );\n\n return result.rows || [];\n } catch (error) {\n console.error(\"Error retrieving previous fun facts:\", error);\n return [];\n }\n}\n\n/**\n * Deletes existing fun facts for the specified dates\n * @param dates Array of date strings in ISO format\n */\nasync function deleteExistingFunFacts(dates) {\n try {\n for (const date of dates) {\n await sqlite.execute(\n `DELETE FROM ${TABLE_NAME}\n WHERE date = ? AND tags = ?`,\n [date, TAG_NAME],\n );\n }\n } catch (error) {\n console.error(\"Error deleting existing fun facts:\", error);\n }\n}\n\n/**\n * Inserts a fun fact into the memories database\n * @param date Date for the fun fact in ISO format\n * @param factText The fun fact text\n */\nasync function insertFunFact(date, factText) {\n try {\n await sqlite.execute(\n `INSERT INTO ${TABLE_NAME} (id, date, text, createdBy, createdDate, tags)\n VALUES (?, ?, ?, ?, ?, ?)`,\n [\n nanoid(10),\n date,\n `fun fact: ${factText}`,\n \"fun_fact_generator\",\n Date.now(),\n TAG_NAME,\n ],\n );\n } catch (error) {\n console.error(`Error inserting fun fact for ${date}:`, error);\n }\n}\n\n/**\n * Generates fun facts for the next 7 days using Claude AI\n * @param previousFacts Previous fun facts to avoid duplication\n * @returns Array of generated fun facts\n */\nasync function generateFunFacts(previousFacts) {\n try {\n // Get API key from environment\n const apiKey = Deno.env.get(\"ANTHROPIC_API_KEY\");\n if (!apiKey) {\n console.error(\"Anthropic API key is not configured.\");\n return null;\n }\n\n // Initialize Anthropic client\n const anthropic = new Anthropic({ apiKey });\n\n // Format previous facts for the prompt\n const previousFactsText = previousFacts\n .map((fact) => `- ${fact.date}: ${fact.text}`)\n .join(\"\\n\");\n\n // Get today's date in US Eastern Time\n const today = DateTime.now().setZone(\"America/New_York\").startOf(\"day\");\n\n // Hana's birthday - adjust as needed\n const hanaBirthday = \"2025-03-26\";\n const currentDate = today.toFormat(\"yyyy-MM-dd\");\n\n // Format dates for the next 7 days\n const nextSevenDays = Array.from({ length: 7 }, (_, i) => {\n const date = today.plus({ days: i });\n return date.toFormat(\"yyyy-MM-dd\");\n });\n\n const message =\n `Generate 7 unique fun facts, one for each of the next 7 days. Each fact should be brief (1-2 sentences) and relate to one of these categories:\n\n1. Infant development appropriate for Miss Hana, who was born on ${hanaBirthday} (today is ${currentDate})\n2. Cute animals: capybaras, border collies, elephants, or penguins\n3. Plants: monsteras, orchids, ficus, or other common houseplants\n\nRequirements:\n- Each fact must be interesting, educational, and appropriate for formal communication\n- Keep each fact to 1-2 sentences maximum\n- Use different categories across the 7 days for variety\n- Ensure facts are scientifically accurate\n- Make infant development facts age-appropriate for Miss Hana based on her birthdate\n- Do not repeat any previous facts from this list:\n\n${previousFactsText}\n\nFORMAT YOUR RESPONSE AS STRUCTURED JSON INSIDE XML TAGS:\n\n[\n {\n \"date\": \"YYYY-MM-DD\",\n \"text\": \"The fun fact text here\",\n \"category\": \"infant\" | \"animal\" | \"plant\"\n },\n ... (repeat for all 7 days)\n]\n\n\nUse these dates for the 7 facts:\n${nextSevenDays.join(\"\\n\")}`;\n\n console.log({ message });\n\n const response = await anthropic.messages.create({\n model: \"claude-3-5-sonnet-latest\",\n max_tokens: 1000,\n messages: [\n {\n role: \"user\",\n content: [\n {\n type: \"text\",\n text: message,\n },\n ],\n },\n ],\n });\n\n // Extract and parse the response\n const responseText = response.content[0].text.trim();\n\n // Extract JSON between tags\n const factsMatch = responseText.match(/([\\s\\S]*?)<\\/facts>/);\n\n if (!factsMatch || factsMatch.length < 2) {\n console.error(\"Failed to parse facts from response:\", responseText);\n return [];\n }\n\n try {\n // Parse the JSON from the extracted content\n const factData = JSON.parse(factsMatch[1]);\n\n // Validate and format facts\n const facts = factData.map((fact) => ({\n date: fact.date,\n text: fact.text,\n category: fact.category,\n }));\n\n return facts;\n } catch (parseError) {\n console.error(\"Failed to parse facts JSON:\", parseError);\n console.log(\"Raw response:\", responseText);\n\n // Fallback to regex parsing if JSON parsing fails\n return parseFallbackFacts(responseText, nextSevenDays);\n }\n } catch (error) {\n console.error(\"Error generating fun facts:\", error);\n return [];\n }\n}\n\n/**\n * Fallback parser for when JSON parsing fails\n * @param responseText The raw response text from Claude\n * @param expectedDates Array of expected dates for facts\n * @returns Array of parsed facts\n */\nfunction parseFallbackFacts(responseText, expectedDates) {\n // Try to extract facts using regex\n const factPattern = /(\\d{4}-\\d{2}-\\d{2})[\"']?[,:]?\\s*[\"']?(.*?)[\"']?[,}]/gs;\n const facts = [];\n let match;\n\n while ((match = factPattern.exec(responseText)) !== null) {\n if (match.length >= 3) {\n facts.push({\n date: match[1],\n text: match[2].trim(),\n });\n }\n }\n\n // If we couldn't extract any facts, try another approach\n if (facts.length === 0) {\n // Look for quoted text that might be facts\n const lines = responseText.split(\"\\n\");\n let currentDate = null;\n\n for (let i = 0; i < lines.length; i++) {\n const line = lines[i].trim();\n\n // Check if this line contains a date\n const dateMatch = line.match(/(\\d{4}-\\d{2}-\\d{2})/);\n if (dateMatch) {\n currentDate = dateMatch[1];\n } else if (currentDate && line.length > 10) {\n // If we have a date and this is a non-empty line, treat it as a fact\n facts.push({\n date: currentDate,\n text: line.replace(/[\"']/g, \"\").trim(),\n });\n currentDate = null;\n }\n }\n }\n\n // As a last resort, assign facts to dates in order\n if (facts.length === 0 && expectedDates.length > 0) {\n // Extract any sentences that look like facts\n const sentences = responseText.match(/[^.!?]+[.!?]+/g) || [];\n\n for (let i = 0; i < Math.min(sentences.length, expectedDates.length); i++) {\n let factText = sentences[i].trim();\n\n // Clean up the fact text\n factText = factText.replace(/[\"']/g, \"\").trim();\n\n if (factText.length > 10) {\n facts.push({\n date: expectedDates[i],\n text: factText,\n });\n }\n }\n }\n\n return facts;\n}\n\n/**\n * Main function to generate and store fun facts for the next 7 days\n */\nexport async function generateAndStoreFunFacts() {\n try {\n // Get previous fun facts\n const previousFacts = await getPreviousFunFacts();\n\n // Generate new fun facts\n const newFacts = await generateFunFacts(previousFacts);\n\n if (!newFacts || newFacts.length === 0) {\n console.error(\"Failed to generate fun facts\");\n return \"Failed to generate fun facts\";\n }\n\n // Get dates for the next 7 days\n const today = DateTime.now().setZone(\"America/New_York\").startOf(\"day\");\n const nextSevenDays = Array.from({ length: 7 }, (_, i) => {\n return today.plus({ days: i }).toFormat(\"yyyy-MM-dd\");\n });\n\n // Delete existing fun facts for these dates\n await deleteExistingFunFacts(nextSevenDays);\n\n // Insert new fun facts\n for (const fact of newFacts) {\n await insertFunFact(fact.date, fact.text);\n }\n\n console.log(`Generated and stored ${newFacts.length} fun facts`);\n return `Generated and stored ${newFacts.length} fun facts`;\n } catch (error) {\n console.error(\"Error in generateAndStoreFunFacts:\", error);\n return \"Error generating fun facts\";\n }\n}\n\n/**\n * Cron job entry point - runs weekly to generate fun facts for the next 7 days\n * Intended to be used as a Val Town cron job\n */\nexport default async function() {\n console.log(\"Running fun facts generation cron job...\");\n return await generateAndStoreFunFacts();\n}\n\n// For testing with Deno directly\nif (import.meta.main) {\n await generateAndStoreFunFacts();\n}",[],[313,337],"4d65b92e-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[243,244,245,246,247],[313],"4d653954-3d2c-11f0-bd13-76b3cceeab13","setupTelegramChatDb.ts","ff5ad839-f0dd-4774-93aa-e7c0a3c3bb6c","356cd243-aa3f-4ff0-b900-5256f26a16e4",["D",1748592093410],"// Script to set up the telegram_chats table in SQLite\n// Run this script manually to create the database table\n\nexport default async function setupTelegramChatDb() {\n try {\n // Import SQLite module\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n // Create the telegram_chats table\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS telegram_chats (\n id INTEGER PRIMARY KEY AUTOINCREMENT,\n chat_id INTEGER NOT NULL,\n sender_id TEXT NOT NULL,\n sender_name TEXT NOT NULL,\n message TEXT NOT NULL,\n timestamp INTEGER NOT NULL,\n is_bot BOOLEAN NOT NULL\n )\n `);\n\n // Create an index on chat_id for faster lookups\n await sqlite.execute(`\n CREATE INDEX IF NOT EXISTS idx_telegram_chats_chat_id ON telegram_chats(chat_id)\n `);\n\n return \"Telegram chat database table created successfully.\";\n } catch (error) {\n console.error(\"Error setting up telegram_chats table:\", error);\n throw error;\n }\n}\n",[],[355,348],"dbUtils","4d6536ac-3d2c-11f0-bd13-76b3cceeab13","populateMemoryIds.ts","0344b7f4-86a0-433e-809c-ab8a2971044d",["D",1748592093410],"// Script to populate random IDs for existing memory rows\n// Run this script manually to generate IDs for rows that don't have them\n\nimport { nanoid } from \"https://esm.sh/nanoid@5.0.5\";\n\nexport default async function populateMemoryIds() {\n try {\n // Import SQLite module\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n // Get all rows without an ID\n const rowsWithoutId = await sqlite.execute(`\n SELECT rowid FROM memories WHERE id IS NULL OR id = ''\n `);\n\n console.log(\n `Found ${rowsWithoutId.rows.length} rows without IDs. Generating IDs...`\n );\n\n // Update each row with a unique ID\n let updatedCount = 0;\n for (const row of rowsWithoutId.rows) {\n const id = nanoid(10); // Generate a 10-character ID\n await sqlite.execute({\n sql: `UPDATE memories SET id = :id, createdDate = :timestamp WHERE rowid = :rowid`,\n args: {\n id,\n timestamp: Date.now(),\n rowid: row.rowid,\n },\n });\n updatedCount++;\n }\n\n console.log(`Successfully updated ${updatedCount} rows with new IDs.`);\n return `Successfully updated ${updatedCount} rows with new IDs.`;\n } catch (error) {\n console.error(\"Error populating memory IDs:\", error);\n throw error;\n }\n}\n\n// Uncomment to run directly\n// populateMemoryIds();\n",[],[355,357],"4d6530e4-3d2c-11f0-bd13-76b3cceeab13","populateCreatedBy.ts","dde3c968-7993-4ea9-8296-6b53d2cf8489",["D",1748592093410],"// Script to populate createdBy for existing memory rows\n// Run this script manually to set createdBy based on memory content\n\nexport default async function populateCreatedBy() {\n try {\n // Import SQLite module\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n // Get all rows without a createdBy\n const rowsWithoutCreatedBy = await sqlite.execute(`\n SELECT rowid, text FROM memories WHERE createdBy IS NULL OR createdBy = ''\n `);\n\n console.log(\n `Found ${rowsWithoutCreatedBy.rows.length} rows without createdBy. Populating...`\n );\n\n // Update each row with the appropriate createdBy\n let updatedCount = 0;\n for (const row of rowsWithoutCreatedBy.rows) {\n let createdBy = \"chat\"; // Default value\n const text = row.text?.toLowerCase() || \"\";\n\n // Determine createdBy based on text content\n if (text.includes(\"weather\")) {\n createdBy = \"weather\";\n } else if (text.includes(\"expecting\")) {\n createdBy = \"usps\";\n }\n\n // Set tags based on createdBy\n let tags = \"\";\n if (createdBy === \"weather\") {\n tags = \"weather\";\n } else if (createdBy === \"usps\") {\n tags = \"mail\";\n }\n\n await sqlite.execute({\n sql: `UPDATE memories\n SET createdBy = :createdBy,\n tags = :tags,\n createdDate = COALESCE(createdDate, :timestamp)\n WHERE rowid = :rowid`,\n args: {\n createdBy,\n tags,\n timestamp: Date.now(),\n rowid: row.rowid,\n },\n });\n updatedCount++;\n }\n\n console.log(\n `Successfully updated ${updatedCount} rows with createdBy values.`\n );\n return `Successfully updated ${updatedCount} rows with createdBy values.`;\n } catch (error) {\n console.error(\"Error populating createdBy:\", error);\n throw error;\n }\n}\n\n// Uncomment to run directly\n// populateCreatedBy();\n",[],[355,364],"4d652e3c-3d2c-11f0-bd13-76b3cceeab13","migrateMemoriesDb.ts","7d7f97c1-a994-4593-8f66-10a87e370be6",["D",1748592093410],"// Script to migrate the memories table in SQLite\n// Run this script manually to add new columns to the memories table\n\nexport default async function migrateMemoriesDb() {\n try {\n // Import SQLite module\n const { sqlite } = await import(\"https://esm.town/v/stevekrouse/sqlite\");\n\n // First, backup the current table contents\n const backup = await sqlite.execute(`SELECT * FROM memories`);\n console.log(\"Backup of memories table (JSON format):\");\n console.log(JSON.stringify(backup.rows, null, 2));\n\n // Create a new table with the updated schema\n await sqlite.execute(`\n CREATE TABLE IF NOT EXISTS memories_new (\n id TEXT,\n date TEXT,\n text TEXT,\n createdBy TEXT,\n createdDate INTEGER,\n tags TEXT\n )\n `);\n\n // Copy data from old table to new table\n await sqlite.execute(`\n INSERT INTO memories_new (date, text)\n SELECT date, text FROM memories\n `);\n\n // Drop the old table\n await sqlite.execute(`DROP TABLE memories`);\n\n // Rename the new table to the original name\n await sqlite.execute(`ALTER TABLE memories_new RENAME TO memories`);\n\n // Create an index on id for faster lookups\n await sqlite.execute(`\n CREATE INDEX IF NOT EXISTS idx_memories_id ON memories(id)\n `);\n\n return \"Memories table migrated successfully with new columns.\";\n } catch (error) {\n console.error(\"Error migrating memories table:\", error);\n throw error;\n }\n}\n\nmigrateMemoriesDb();\n",[],[355,371],"4d653d0a-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[238,239,240,241],[355],"4d6529b4-3d2c-11f0-bd13-76b3cceeab13","2227b3b1-7343-4810-9e0b-ba11187160c3","cc9ede02-c8b3-4665-a1da-b8c94bd5f114",["D",1748592093410],"# React Hono Val Town Project Starter Template\n\nThis is a starter template for a full-stack app in a Val Town Project. The app itself is a simple persistent message board.\n\nThis app is broken up into three folders, corresponding to where the code in those folders run:\n\n- `backend/` - runs on Val Town serverless infrastructure in Deno\n- `frontend/` - runs in the user's browser\n- `shared/` - runs in both the frontend and backend\n\nThe entrypoint of this app is `backend/index.ts`, which is the Hono HTTP server, which serves the HTML, CSS, and JS, which run client-side.\n\nYou can click into each of those folders to view their respective READMEs.",[],[186,278],"4d6524dc-3d2c-11f0-bd13-76b3cceeab13","utils.ts","cbc3bc8a-9601-483f-81c4-c32e7868e78c","010e78fa-da14-4ad1-a3d9-7aa581193690",["D",1748592093410],"export interface Message {\n content: string;\n}\n\nexport const MESSAGE_LIMIT = 5;",[],[186,299,389],"4d652202-3d2c-11f0-bd13-76b3cceeab13","f70ba18a-55e9-41ee-a8bd-51fc369a1c5c",["D",1748592093410],"/**\n * @jsxImportSource https://esm.sh/react@18.2.0\n */\n\nexport interface Memory {\n id: string; // UUID\n date: string; // ISO 8601 date string or similar\n text: string;\n createdBy?: string | null; // Optional: Who/what created the memory\n createdDate?: number | null; // Optional: Unix timestamp (integer)\n tags?: string | null; // Optional: Comma-separated string or JSON\n}\n",[],[186,299,292],"4d651f96-3d2c-11f0-bd13-76b3cceeab13","b513d0d0-7657-4e42-ad29-056a45b17a15",["D",1748592093410],"# Shared\n\nIt's common to have code and types that are needed on both the frontend and the backend. It's important that you write this code in a particularly defensive way because it's limited by what both environments support:\n\n![shapes at 25-02-25 11.57.13.png](https://imagedelivery.net/iHX6Ovru0O7AjmyT5yZRoA/75db1d51-d9b3-45e0-d178-25d886c10700/public)\n\nFor example, you *cannot* use the `Deno` keyword. For imports, you can't use `npm:` specifiers, so we reccomend `https://esm.sh` because it works on the server & client. You *can* use TypeScript because that is transpiled in `/backend/index.ts` for the frontend. Most code that works on the frontend tends to work in Deno, because Deno is designed to support \"web-standards\", but there are definitely edge cases to look out for.",[],[186,299,278],"4d6526d0-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[233,234,235],[186,299],"4d651b22-3d2c-11f0-bd13-76b3cceeab13","index.tsx","03dc779e-cb00-4670-be73-27bb907a1577",["D",1748592093410],"/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport { createRoot } from \"https://esm.sh/react-dom@18.2.0/client\";\nimport { App } from \"./components/App.tsx\";\n\nconst rootElement = document.getElementById(\"root\");\nif (!rootElement) {\n throw new Error(\"Root element #root not found\");\n}\n\nconst root = createRoot(rootElement);\nroot.render();\n",[],[186,187,413],"4d651898-3d2c-11f0-bd13-76b3cceeab13","index.html","4edee998-2935-4c49-a14f-e2b78ccf782a",["D",1748592093410],"\n\n \n \n \n Stevens' Notebook\n \n \n \n \n \n \n \n \n \n

\n \n \n \n\n",[],[186,187,420],"4d651492-3d2c-11f0-bd13-76b3cceeab13","favicon.svg","59a24e94-4a72-408e-b416-beb7a228917e",["D",1748592093410],"",[],[186,187,427],"4d651212-3d2c-11f0-bd13-76b3cceeab13","af073d85-b9a9-46c3-bc64-54a2763ff9a5",["D",1748592093410],"# Frontend\n\nThis template is a classic client-side-only React app. \n\n## `index.html`\n\nThe **entrypoint** for the frontend is `/frontend/index.html`. This is the HTML page that is returned at the root from `/backend/index.ts`.\n\nThis HTML file imports `/frontend/style.css` from `/public/style.css` and `/frontend/favicon.svg` from `/frontend/favicon.svg`. Everything in `/frontend/` is mapped to `/public` by `/backend/index.ts`. This is just a convention. You could import & serve everything out of the same folder name.\n\nThis HTML file has a `

`, which is where we mount the React app.\n\nThis HTML file imports `/frontend/index.tsx` from `/public/index.tsx`, which is the **entrypoint** for all frontend JavaScript, including all the React. It is not a problem that it imports a file with a `.tsx` extension becaues browsers ignore file extensions. They only pay attention to content-types, which is great, because all these files will be returned as transpiled JS with the appropriate JS content type by [stevekrouse/utils/serve-public](https://www.val.town/x/stevekrouse/utils/branch/main/code/serve-public/README.md).)\n\n## `index.tsx`\n\nThis file is the **entrypoint** for frontend JavaScript. It imports the React app from `/frontend/components/App.tsx` and mounts it on `

`.\n\nIt also looks for *bootstrapped* JSON data at `window.__INITIAL_DATA`, and passes that only to the ``.\n\n## `favicon.svg`\n\nAs of this writing Val Town only supports text files, which is why the favicon is an SVG and not an .ico or any other binary image format. If you need binary file storage, check out [Blob Storage](https://docs.val.town/std/blob/).\n\n## `components/`\n\nThis directory is where the React components are stored. They're pretty standard client-side React components.",[],[186,187,278],["D",1748592093410],[],[186,187,94,87],["D",1748592093410],[],[186,187,94,92],["D",1748592093410],[],[186,187,94,75],["D",1748592093410],[225,226,227],[186,187,94],"4d651d5c-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[224,228,229,230,231],[186,187],"4d648ae0-3d2c-11f0-bd13-76b3cceeab13","index.ts","137b348b-288e-4a2b-8052-c4cac32ab17b","8f7ee81b-1672-40e5-b545-24e490109f3a",["D",1748592093410],"/**\n * @jsxImportSource https://esm.sh/react@18.2.0\n */\nimport { blob } from \"https://esm.town/v/std/blob\";\nimport { readFile, serveFile } from \"https://esm.town/v/std/utils@85-main/index.ts\";\nimport { Hono } from \"npm:hono\";\nimport { type Memory } from \"../shared/types.ts\";\nimport { createMemory, deleteMemory, getAllMemories, updateMemory } from \"./database/queries.ts\";\n\nconst app = new Hono();\n\n// Unwrap Hono errors to see original error details\napp.onError((err, c) => {\n console.error(\"Error in Hono route:\", err);\n // Optionally re-throw or return a custom error response\n // throw err;\n return c.json({ error: \"Internal Server Error\", details: err.message }, 500);\n});\n\n// --- API Routes for Memories ---\n\n// GET /api/memories - Retrieve all memories\napp.get(\"/api/memories\", async (c) => {\n const memories = await getAllMemories();\n return c.json(memories);\n});\n\n// POST /api/memories - Create a new memory\napp.post(\"/api/memories\", async (c) => {\n const body = await c.req.json>();\n if (!body.text) {\n return c.json({ error: \"Memory text cannot be empty\" }, 400);\n }\n const newMemory = await createMemory(body);\n return c.json(newMemory, 201);\n});\n\n// PUT /api/memories/:id - Update an existing memory\napp.put(\"/api/memories/:id\", async (c) => {\n const id = c.req.param(\"id\");\n const body = await c.req.json>>();\n\n if (Object.keys(body).length === 0) {\n return c.json({ error: \"No update fields provided\" }, 400);\n }\n\n try {\n await updateMemory(id, body);\n return c.json({ success: true });\n } catch (error) {\n // Handle cases like memory not found, if needed\n console.error(\"Error updating memory:\", error);\n return c.json(\n { error: \"Failed to update memory\", details: error.message },\n 500,\n );\n }\n});\n\n// DELETE /api/memories/:id - Delete a memory\napp.delete(\"/api/memories/:id\", async (c) => {\n const id = c.req.param(\"id\");\n try {\n await deleteMemory(id);\n return c.json({ success: true });\n } catch (error) {\n console.error(\"Error deleting memory:\", error);\n return c.json(\n { error: \"Failed to delete memory\", details: error.message },\n 500,\n );\n }\n});\n\n// --- Blob Image Serving Routes ---\n\n// GET /api/images/:filename - Serve images from blob storage\napp.get(\"/api/images/:filename\", async (c) => {\n const filename = c.req.param(\"filename\");\n\n try {\n // Get image data from blob storage\n const imageData = await blob.get(filename);\n\n if (!imageData) {\n return c.json({ error: \"Image not found\" }, 404);\n }\n\n // Determine content type based on file extension\n let contentType = \"application/octet-stream\"; // Default\n if (filename.endsWith(\".jpg\") || filename.endsWith(\".jpeg\")) {\n contentType = \"image/jpeg\";\n } else if (filename.endsWith(\".png\")) {\n contentType = \"image/png\";\n } else if (filename.endsWith(\".gif\")) {\n contentType = \"image/gif\";\n } else if (filename.endsWith(\".svg\")) {\n contentType = \"image/svg+xml\";\n }\n\n // Return the image with appropriate headers\n return new Response(imageData, {\n headers: {\n \"Content-Type\": contentType,\n \"Cache-Control\": \"public, max-age=3600\", // Cache for 1 hour\n },\n });\n } catch (error) {\n console.error(`Error serving image ${filename}:`, error);\n return c.json(\n { error: \"Failed to load image\", details: error.message },\n 500,\n );\n }\n});\n\n// --- Static File Serving ---\n\n// Serve index.html at the root /\napp.get(\"/\", async (c) => {\n let html = await readFile(\"/dashboard/frontend/index.html\", import.meta.url);\n // Inject view source link\n const viewSourceLink = import.meta.url.replace(\"esm.town\", \"val.town\");\n const viewSourceTag =\n `View Source`;\n html = html.replace(\"\", `${viewSourceTag}`);\n return c.html(html);\n});\n\n// Serve all /frontend files (except index.html) under /public\napp.get(\"/public/*\", (c) =>\n serveFile(\n c.req.path.replace(\"/public\", \"/dashboard/frontend\"),\n import.meta.url,\n ));\n\n// HTTP vals expect an exported \"fetch handler\"\nexport default app.fetch;",[],[186,463,456],"backend","4d6485a4-3d2c-11f0-bd13-76b3cceeab13","aad86598-8de1-4ddf-81f1-a299a499ac02",["D",1748592093410],"# Backend\n\nThis folder contains all backend-only code in this project. It's broken up into:\n\n* `index.ts` - this is the **entrypoint** for this whole project\n* `database/` - this contains the code for interfacing with the app's SQLite database table\n\n## Hono\n\nThis app uses [Hono](https://hono.dev/) as the API framework. You can think of Hono as a replacement for [ExpressJS](https://expressjs.com/) that works in serverless environments like Val Town or Cloudflare Workers. If you come from Python or Ruby, Hono is also a lot like [Flask](https://github.com/pallets/flask) or [Sinatra](https://github.com/sinatra/sinatra), respectively.\n\n## Serving assets to the frontend\n\nThis backend HTTP server is responsible for serving all static assets to the browser to render the app, including HTML, JavaScript (including all client-side React), CSS, and even the favicon SVG.\n\nIn a normal server environment, you would likely use a middleware [like this one](https://hono.dev/docs/getting-started/nodejs#serve-static-files) to serve static files. Some frameworks or deployment platforms automatically make any content inside a `public/` folder public. \n\nHowever in Val Town you need to handle this yourself, and it can be suprisingly difficult to read and serve files in a Val Town Project. This template uses helper functions from [stevekrouse/utils/serve-public](https://www.val.town/x/stevekrouse/utils/branch/main/code/serve-public/README.md), which handle reading project files in a way that will work across branches and forks, automatically transpiles typescript to javascript, and assigns content-types based on the file's extension.\n\n### `index.html`\n\nThe most complicated part of this backend API is serving index.html. In this app (like most apps) we serve it at the root, ie `GET /`. \n\nWe *bootstrap* `index.html` with some initial data from the server, so that it gets dynamically injected JSON data without having to make another round-trip request to the server to get that data on the frontend. This is a common pattern for client-side rendered apps.\n\n## CRUD API Routes\n\nThis app has two CRUD API routes: for reading and inserting into the messages table. They both speak JSON, which is standard. They import their functions from `/backend/database/queries.ts`. These routes are called from the React app to refresh and update data.\n\n## Errors\n\nHono and other API frameworks have a habit of swallowing up Errors. We turn off this default behavior by re-throwing errors, because we think most of the time you'll want to see the full stack trace instead of merely \"Internal Server Error\". You can customize how you want errors to appear.",[],[186,463,278],"4d647fdc-3d2c-11f0-bd13-76b3cceeab13","queries.ts","bbe21360-acb5-4a58-85aa-ffba4c3a47e7","7efbb13b-cd48-4145-90f6-657ccd3d039a",["D",1748592093410],"import { sqlite } from \"https://esm.town/v/stevekrouse/sqlite\";\nimport { type Memory } from \"../../shared/types.ts\";\n\n// toggle the commented out table name to switch between the demo and production tables\n// const tableName = \"memories\";\nconst tableName = \"memories_demo\";\n\nexport async function getAllMemories(): Promise {\n const result = await sqlite.execute(\n `SELECT id, date, text, createdBy, createdDate, tags FROM ${tableName}\n ORDER BY createdDate DESC`\n );\n\n // Ensure correct typing, especially for potential null values\n return result.rows.map((row) => ({\n id: row.id as string,\n date: row.date as string | null,\n text: row.text as string,\n createdBy: row.createdBy as string | null,\n createdDate: row.createdDate as number | null,\n tags: row.tags as string | null,\n })) as Memory[];\n}\n\nexport async function createMemory(\n memory: Omit\n): Promise {\n // Use nanoid like in getWeather.ts\n const { nanoid } = await import(\"https://esm.sh/nanoid@5.0.5\");\n const newId = nanoid(10); // Generate a shorter ID like in getWeather\n const createdDate = memory.createdDate ?? Date.now(); // Default to now if not provided\n\n await sqlite.execute(\n `INSERT INTO ${tableName} (id, date, text, createdBy, createdDate, tags)\n VALUES (?, ?, ?, ?, ?, ?)`,\n [\n newId,\n memory.date,\n memory.text,\n memory.createdBy || \"dashboard\", // Default creator\n createdDate,\n memory.tags,\n ]\n );\n\n return {\n ...memory,\n id: newId,\n createdDate,\n };\n}\n\nexport async function updateMemory(\n id: string,\n memory: Partial>\n): Promise {\n const fields = Object.keys(memory).filter((key) => key !== \"id\");\n const values = Object.values(memory);\n\n if (fields.length === 0) {\n throw new Error(\"No fields provided for update\");\n }\n\n const setClause = fields.map((field) => `${field} = ?`).join(\", \");\n\n await sqlite.execute(`UPDATE ${tableName} SET ${setClause} WHERE id = ?`, [\n ...values,\n id,\n ]);\n}\n\nexport async function deleteMemory(id: string): Promise {\n await sqlite.execute(`DELETE FROM ${tableName} WHERE id = ?`, [id]);\n}\n",[],[186,463,478,471],"database","4d647d8e-3d2c-11f0-bd13-76b3cceeab13","7146785f-ca49-4088-8622-e2e9401c6748",["D",1748592093410],"# Database\n\nThis app uses [Val Town SQLite](https://docs.val.town/std/sqlite/) to manage data. Every Val Town account comes with a free SQLite database, hosted on [Turso](https://turso.tech/). This folder is broken up into two files:\n\n* `migrations.ts` - code to set up the database tables the app needs\n* `queries.ts` - functions to run queries against those tables, which are imported and used in the main Hono server in `/backend/index.ts`\n\n## Migrations\n\nIn `backend/database/migrations.ts`, this app creates a new SQLite table `reactHonoStarter_messages` to store messages. \n\nThis \"migration\" runs once on every app startup because it's imported in `index.ts`. You can comment this line out for a slight (30ms) performance improvement on cold starts. It's left in so that users who fork this project will have the migration run correctly.\n\nSQLite has much more limited support for altering existing tables as compared to other databases. Often it's easier to create new tables with the schema you want, and then copy the data over. Happily LLMs are quite good at those sort of database operations, but please reach out in the [Val Town Discord](https://discord.com/invite/dHv45uN5RY) if you need help.\n\n## Queries\n\nThe queries file is where running the migrations happen in this app. It'd also be reasonable for that to happen in index.ts, or as is said above, for that line to be commented out, and only run when actual changes are made to your database schema.\n\nThe queries file exports functions to get and write data. It relies on shared types and data imported from the `/shared` directory.",[],[186,463,478,278],"4d6483ba-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[219,220],[186,463,478],"4d649148-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[218,221,222],[186,463],"4d652be4-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[217,223,232,236],[186],"4d647546-3d2c-11f0-bd13-76b3cceeab13","testDailyBrief.ts","eb1e75f2-8f64-4af9-9c8c-575611bd37f1","b774b9f0-98b6-4e30-8bff-6e3ff1767a90",["D",1748592093410],"// Send a test of the daily brief to a test chat id\n\nimport { sendDailyBriefing } from \"./sendDailyBrief.ts\";\nimport { DateTime } from \"https://esm.sh/luxon@3.4.4\";\n\nexport async function testDailyBrief() {\n try {\n const testChatId = Deno.env.get(\"TEST_TELEGRAM_CHAT_ID\");\n if (!testChatId) {\n console.error(\"TEST_TELEGRAM_CHAT_ID is not configured.\");\n return;\n }\n\n console.log(\"Sending test daily briefing for tomorrow...\");\n const result = await sendDailyBriefing(\n testChatId,\n DateTime.now()\n .setZone(\"America/New_York\")\n .startOf(\"day\")\n .plus({ days: 1 })\n );\n return result;\n } catch (error) {\n console.error(\"Error sending test daily briefing:\", error);\n throw error;\n }\n}\n\n// Export as default for direct execution\ntestDailyBrief();\n",[],[505,498],"dailyBriefing","4d645dd6-3d2c-11f0-bd13-76b3cceeab13","sendDailyBrief.ts","ebd88588-f227-4ec0-ac67-bd737c3b9409",["D",1748592093410],"import { Bot } from \"https://deno.land/x/grammy@v1.35.0/mod.ts\";\nimport { DateTime } from \"https://esm.sh/luxon@3.4.4\";\nimport Anthropic from \"npm:@anthropic-ai/sdk@0.24.3\";\nimport { backstory } from \"../backstory.ts\";\nimport {\n BOT_SENDER_ID,\n BOT_SENDER_NAME,\n storeChatMessage,\n} from \"../importers/handleTelegramMessage.ts\";\nimport {\n formatMemoriesForPrompt,\n getRelevantMemories,\n} from \"../memoryUtils.ts\";\n\nasync function generateBriefingContent(anthropic, memories, today, isSunday) {\n try {\n const weekdaysHelp = generateWeekDays(today);\n // Create system prompt\n const systemPrompt = `${backstory}`;\n\n // Create user message\n const userMessage = {\n role: \"user\",\n content: `Today it is your duty to provide a daily briefing summarizing important information for the day. The briefing should have the following sections:\n\nBegin with a formal morning greeting, maintaining professional decorum. Try to mix up the greetings, for example mention the season or the weather.\n\n*Today*\n\nNote any reminders about today's affairs.\nProvide a summary of today's meteorological conditions.\nDetail the day's postal correspondence, highlighting any significant items such as important documents, personal letters, or parcels. Advertisements need not be mentioned. If there is no mail, this section may be omitted.\n\n*Looking Ahead*\n\nOffer a brief overview of forthcoming engagements and tasks for the remainder of the week, with particular attention to tomorrow's schedule.\nShould there be noteworthy meteorological conditions anticipated later in the week (such as precipitation or significant temperature variations), these should be mentioned. If the weather is unremarkable, this need not be addressed.\nOne concise paragraph, 2-3 sentences maximum, without bullet points or subsections.\n\n*Daily fact*\n\nInclude the fun fact for today from the memories. This will be labeled with \"fun fact:\" in the text field.\nIf no fun fact is available for today, you may omit this section.\n\nSign off with a formal greeting.\n\nUse the following memories to fill in the information for your briefing:\n\n${formatMemoriesForPrompt(memories)}\n\nResponse guidelines:\n\nAlways follow these rules:\n- Use Telegram-friendly markdown format (supports *bold*, _italic_, [links](http://example.com))). Do not use markdown headings like ## as they are not supported in Telegram messages.\n- Make the briefing easily skimmable by using clear sections. Use bolded text to begin each section, eg *Today*\n- Use emojis to help reinforce the content visually. Use emojis for specific concepts like sun/rainy weather, paper forms for a logistical todo, etc. Don't use emojis for general concepts like \"today\"\n\n- Address the message to \"Sir and Madam\".\n- Maintain a formal and professional tone throughout.\n- Use phrases characteristic of Stevens' speech patterns, such as:\n - \"I should say...\"\n - \"I would venture...\"\n - \"If I may be so bold...\"\n - \"It would appear that...\"\n - \"One might observe...\"\n - \"I trust you will find...\"\n - \"I would be remiss not to mention...\"\n- Avoid contractions (use \"do not\" instead of \"don't\")\n- Express opinions tentatively and with great deference\n- Use British English spelling and terminology\n- Keep the content concise but informative, maintaining the highest standards of professional communication\n- You should reference upcoming days as \"today\", \"tomorrow\", \"Thursday\", etc. rather than using dates. Here's a guide:\n${weekdaysHelp}`,\n };\n\n console.log(\"Sending prompt to Anthropic...\", userMessage);\n\n const response = await anthropic.messages.create({\n model: \"claude-3-7-sonnet-latest\",\n max_tokens: 30000,\n system: systemPrompt,\n thinking: {\n type: \"enabled\",\n budget_tokens: 24000,\n },\n messages: [userMessage],\n });\n\n const text = response.content.find((block) => block.type === \"text\").text;\n\n return text;\n } catch (error) {\n console.error(\"Briefing generation error:\", error);\n throw error;\n }\n}\n\nexport async function sendDailyBriefing(chatId?: string, today?: DateTime) {\n // Get API keys from environment\n const apiKey = Deno.env.get(\"ANTHROPIC_API_KEY\");\n const telegramToken = Deno.env.get(\"TELEGRAM_TOKEN\");\n\n // If no chatId is provided, try to get it from environment variables\n if (!chatId) {\n chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n }\n\n if (!apiKey) {\n console.error(\"Anthropic API key is not configured.\");\n return;\n }\n\n if (!telegramToken) {\n console.error(\"TELEGRAM_TOKEN is not configured.\");\n return;\n }\n\n if (!chatId) {\n console.error(\"No chat ID provided or found in environment variables.\");\n return;\n }\n\n // Initialize Anthropic client\n const anthropic = new Anthropic({ apiKey });\n\n // Initialize Telegram bot\n const bot = new Bot(telegramToken);\n\n // Use provided date or get today's date in US Eastern Time\n if (!today) {\n today = DateTime.now().setZone(\"America/New_York\").startOf(\"day\");\n }\n\n // Get the date of the last Sunday\n const lastSunday = today.startOf(\"week\").minus({ days: 1 });\n\n // Fetch relevant memories using the utility function\n const memories = await getRelevantMemories();\n\n console.log(memories);\n\n // Check if today is Sunday (1 is Sunday in Luxon)\n const isSunday = today.weekday === 7;\n\n // Generate briefing content\n const content = await generateBriefingContent(\n anthropic,\n memories,\n today,\n isSunday\n );\n\n console.log(\"content\", content);\n\n console.log(\"Sending Telegram message...\");\n\n // Format message for Telegram\n // Telegram supports Markdown V2, but it's more restrictive than regular Markdown\n // For simplicity, we'll use the content as is, which should work with basic formatting\n\n // First send the title as a separate message\n\n // disabled title for now, it seemes unnecessary...\n // await bot.api.sendMessage(chatId, `*${title}*`, { parse_mode: \"Markdown\" });\n\n // Then send the main content\n // Telegram has a 4096 character limit per message, so we might need to split it\n const MAX_LENGTH = 4000; // A bit less than 4096 to be safe\n\n if (content.length <= MAX_LENGTH) {\n await bot.api.sendMessage(chatId, content, { parse_mode: \"Markdown\" });\n // Store the briefing in chat history\n await storeChatMessage(\n chatId,\n BOT_SENDER_ID,\n BOT_SENDER_NAME,\n content,\n true\n );\n } else {\n // Split the content into chunks\n const chunks = [];\n let currentChunk = \"\";\n\n for (const line of content.split(\"\\n\")) {\n if ((currentChunk + line + \"\\n\").length > MAX_LENGTH) {\n chunks.push(currentChunk);\n currentChunk = line + \"\\n\";\n } else {\n currentChunk += line + \"\\n\";\n }\n }\n\n if (currentChunk) {\n chunks.push(currentChunk);\n }\n\n // Send each chunk as a separate message and store in chat history\n for (const chunk of chunks) {\n await bot.api.sendMessage(chatId, chunk, { parse_mode: \"Markdown\" });\n // Store each chunk in chat history\n await storeChatMessage(\n chatId,\n BOT_SENDER_ID,\n BOT_SENDER_NAME,\n chunk,\n true\n );\n // Add a small delay to avoid rate limiting\n await new Promise((resolve) => setTimeout(resolve, 500));\n }\n }\n\n console.log(\"Daily briefing sent successfully via Telegram.\");\n return \"Daily briefing sent successfully.\";\n}\n\nfunction generateWeekDays(today) {\n let output = [];\n\n // Add today\n output.push(`* Today: ${today.toFormat(\"EEEE, MMMM d\")}`);\n\n // Add tomorrow\n const tomorrow = today.plus({ days: 1 });\n output.push(`* Tomorrow: ${tomorrow.toFormat(\"EEEE, MMMM d\")}`);\n\n // Add rest of the week\n for (let i = 2; i < 7; i++) {\n const futureDay = today.plus({ days: i });\n output.push(`* ${futureDay.toFormat(\"EEEE, MMMM d\")}`);\n }\n\n return output.join(\"\\n\");\n}\n\n// Example usage:\n// const weekDays = generateWeekDays();\n// console.log(weekDays);\n\n// Export a function that calls sendDailyBriefing with no parameters\n// This maintains backward compatibility with existing cron jobs\nexport default async function (overrideToday?: DateTime) {\n return await sendDailyBriefing(undefined, overrideToday);\n}\n",[],[505,507],"4d642e10-3d2c-11f0-bd13-76b3cceeab13","cronDailyBrief.ts","a89898fa-1835-4ead-b1dc-a738c79dda94",["D",1748592093410],"import { sendDailyBriefing } from \"./sendDailyBrief.ts\";\n\nexport async function cronDailyBrief() {\n try {\n const chatId = Deno.env.get(\"TELEGRAM_CHAT_ID\");\n if (!chatId) {\n console.error(\"TELEGRAM_CHAT_ID is not configured.\");\n return;\n }\n\n console.log(\"Sending scheduled daily briefing...\");\n const result = await sendDailyBriefing(chatId);\n return result;\n } catch (error) {\n console.error(\"Error sending scheduled daily briefing:\", error);\n throw error;\n }\n}\n\n// Export as default for cron scheduling\nexport default cronDailyBrief;\n",[],[505,514],"4d647afa-3d2c-11f0-bd13-76b3cceeab13",["D",1748592093410],[213,214,215],[505],["D",1748592093410],{"_11":12,"_144":81},{"_11":12,"_140":-5,"_80":81,"_31":79,"_141":142,"_143":30},{"_39":40,"_13":16,"_33":36,"_130":97,"_131":70,"_132":71,"_133":101,"_134":135,"_136":137,"_138":14,"_139":12},"ctx",{"_52":-5,"_165":561,"_562":563,"_564":565,"_566":30,"_567":30,"_568":30,"_569":24,"_570":-5,"_571":-7,"_572":-7,"_573":26,"_574":30,"_575":30,"_576":-5,"_577":30,"_578":30,"_579":30,"_580":-5},"ENV",{"_532":533,"_534":535,"_536":16,"_537":538,"_539":540,"_541":542,"_543":544,"_545":546,"_547":548,"_549":550,"_551":552,"_553":554,"_555":556,"_557":558,"_559":560},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","71c1b20804b3232a7e513be260cfffc1932a9e97","RENDER_GIT_BRANCH","POSTHOG_API_TOKEN","phc_nmoyuAkStEZuUp69rYlg1DZgrIt19Gx0Y4Hwf8tGgJP","POSTHOG_API_HOST","https://wart.val.town","WEB_API_TEMPLATE","https://{val}.web.val.run","APP_URL","https://www.val.town","MODULE_URL","https://esm.town","VANITY_WEB_API_TEMPLATE","https://{label}.val.run/","DENOLS_URL","https://lsp.val.town","NODE_ENV","production","CLOUDFLARE_TURNSTILE_SITE_KEY","0x4AAAAAACtaIZJH7s9rBMZ6","GOOGLE_CLIENT_ID","875989696899-jtsqkhlbdv5jrlsk8631p7ipjg3c6o62.apps.googleusercontent.com","STRIPE_PUBLISHABLE_KEY","pk_live_51N2zvWEpZTTLipIkVEwXrUG9AT0yeOOEYxszarZT1jpsZV58Ly6wlg3SltZpHN26CUalOBmUSY7adQRQiSkCZz2500qSXNkf8Y","API_URL","https://api.val.town",[],"preferencesKeybindings","default","preferencesTheme","light","preferencesFixers","preferencesCopilot","preferencesSelectionAi","preferencesCloseBrackets","preferencesChatPanelState","preferencesDprint","preferencesDenofmt","notificationCount","flagDenols","isBusinessOrgMember","businessOrgHandle","isBusinessOrgAdmin","isHigherTownieTier","navCollapsed","currentOrgId"]

History · components