GitHub Release Notes Generator

Generate user-friendly release notes from your GitHub commits

\n\n {isSharedView && (\n

\n Shared View: \n You're viewing shared release notes for {owner}/{repo}.\n

\n )}\n\n {error && (\n

\n Error: \n {error}\n \n

Common issues:

Make sure the repository exists and is spelled correctly
Verify your token has the repo scope
Check that you have access to the repository
Ensure the date range is valid

\n )}\n\n {!isSharedView && (\n

1. Connect to GitHub

\n \n

\n GitHub Personal Access Token\n setToken(e.target.value)}\n className=\"w-full p-2 border rounded\"\n placeholder=\"ghp_...\"\n />\n

\n Needs repo access permissions. \n \n Generate a token\n \n \n Make sure to include repo scope when creating your token.\n \n

\n\n \n\n \n {loading.commits ? 'Loading...' : 'Fetch Commits'}\n \n

\n )}\n\n {!isSharedView && commits.length > 0 && (\n

2. Review Commits

\n \n \n \n {loading.notes ? 'Generating...' : 'Generate Release Notes'}\n \n

\n )}\n\n {releaseNotes && (\n

\n {isSharedView ? 'Shared Release Notes' : '3. Release Notes'}\n

\n \n \n {!isSharedView && (\n

\n \n Copy to Clipboard\n \n

\n )}\n

\n )}\n\n \n

\n );\n}",[],[147,204],"ff5eeefe-318b-11f0-a3c3-569c3dd06744","index.html","6641411e-2c1c-11f0-bfe7-569c3dd06744",68,["D",1747313780300],6,"\n\n \n \n \n GitHub Release Notes Generator\n \n \n \n \n \n

\n\n \n \n \n",[],[147,211],"7c0f2312-2c1c-11f0-99b5-569c3dd06744","RepoSelector.tsx","7c0f23bc-2c1c-11f0-99b5-569c3dd06744","7e53b8df-bba1-4a86-a0cc-a8113885d62b",9,["D",1746716130048],"/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport React from \"https://esm.sh/react@18.2.0\";\n\ninterface RepoSelectorProps {\n owner: string;\n repo: string;\n startDate: string;\n endDate: string;\n onOwnerChange: (owner: string) => void;\n onRepoChange: (repo: string) => void;\n onStartDateChange: (date: string) => void;\n onEndDateChange: (date: string) => void;\n}\n\nexport function RepoSelector({\n owner,\n repo,\n startDate,\n endDate,\n onOwnerChange,\n onRepoChange,\n onStartDateChange,\n onEndDateChange\n}: RepoSelectorProps) {\n // Default to 30 days ago for start date if not set\n React.useEffect(() => {\n if (!startDate) {\n const thirtyDaysAgo = new Date();\n thirtyDaysAgo.setDate(thirtyDaysAgo.getDate() - 30);\n onStartDateChange(thirtyDaysAgo.toISOString().split('T')[0]);\n }\n \n if (!endDate) {\n const today = new Date();\n onEndDateChange(today.toISOString().split('T')[0]);\n }\n }, []);\n\n return (\n

\n Repository Owner\n onOwnerChange(e.target.value)}\n className=\"w-full p-2 border rounded\"\n placeholder=\"e.g., facebook\"\n />\n

\n \n

\n Repository Name\n onRepoChange(e.target.value)}\n className=\"w-full p-2 border rounded\"\n placeholder=\"e.g., react\"\n />\n

\n \n

\n Start Date\n onStartDateChange(e.target.value)}\n className=\"w-full p-2 border rounded\"\n />\n

\n \n

\n End Date\n onEndDateChange(e.target.value)}\n className=\"w-full p-2 border rounded\"\n />\n

\n );\n}",[],[147,228,220],"components","9e409cd6-2cdf-11f0-967c-569c3dd06744","ReleaseNotes.tsx","8cd1b49e-2c1c-11f0-b831-569c3dd06744",64,["D",1746799939267],10,"/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport { marked } from \"https://esm.sh/marked@9.1.5\";\nimport React, { useEffect, useState } from \"https://esm.sh/react@18.2.0\";\n\ninterface ReleaseNotesProps {\n markdown: string;\n onMarkdownChange?: (markdown: string) => void;\n owner?: string;\n repo?: string;\n startDate?: string;\n endDate?: string;\n}\n\ntype SortOption = \"category\" | \"impact\" | \"chronological\";\n\nexport function ReleaseNotes({\n markdown,\n onMarkdownChange,\n owner,\n repo,\n startDate,\n endDate,\n}: ReleaseNotesProps) {\n const [isEditing, setIsEditing] = useState(false);\n const [editableMarkdown, setEditableMarkdown] = useState(markdown);\n const [showShareModal, setShowShareModal] = useState(false);\n const [shareableUrl, setShareableUrl] = useState(\"\");\n const [shareCopied, setShareCopied] = useState(false);\n const [isSharing, setIsSharing] = useState(false);\n const [sortOption, setSortOption] = useState(\"category\");\n const [showPRLinks, setShowPRLinks] = useState(false);\n\n // Update editable markdown when the prop changes\n useEffect(() => {\n setEditableMarkdown(markdown);\n }, [markdown]);\n\n // Configure marked for proper rendering\n // Docs: https://marked.js.org/\n marked.setOptions({\n gfm: true, // GitHub Flavored Markdown\n breaks: true, // Convert \\n to
\n headerIds: true, // Add IDs to headers\n mangle: false, // Don't escape HTML\n pedantic: false, // Don't be overly conformant\n smartLists: true, // Use smarter list behavior\n smartypants: true, // Use \"smart\" typographic punctuation\n });\n\n // Helper function to clean impact indicators from items\n const cleanImpactIndicators = (item: string): string => {\n return item.replace(/\\*\\*(high|medium|low) impact\\*\\*/gi, \"\").trim();\n };\n\n // Sort the markdown content based on the selected option\n const sortMarkdown = (markdown: string, sortOption: SortOption): string => {\n try {\n // Extract sections\n const headerMatch = markdown.match(/^## .*?\\n/);\n const header = headerMatch ? headerMatch[0] : \"## Release Notes\\n\\n\";\n\n // Extract all items with their categories\n const itemRegex = /### (Features|Improvements|Bug Fixes|Other Changes)\\n((?:- \\*\\*.*?\\*\\*: .*?\\n)+)/g;\n const items: { category: string; content: string; items: string[] }[] = [];\n let match;\n\n while ((match = itemRegex.exec(markdown)) !== null) {\n const category = match[1];\n const content = match[2];\n\n // Extract individual items\n const individualItems = content.match(/- \\*\\*.*?\\*\\*: .*?\\n/g) || [];\n\n items.push({\n category,\n content,\n items: individualItems,\n });\n }\n\n if (sortOption === \"category\") {\n // For category sorting, keep the original structure but clean impact indicators\n let categoryMarkdown = header + \"\\n\";\n \n items.forEach(section => {\n categoryMarkdown += `### ${section.category}\\n`;\n \n section.items.forEach(item => {\n categoryMarkdown += cleanImpactIndicators(item) + \"\\n\";\n });\n \n categoryMarkdown += \"\\n\";\n });\n \n return categoryMarkdown;\n }\n\n if (sortOption === \"impact\") {\n // Extract all items with their impact levels\n const allItems: { item: string; impact: number }[] = [];\n\n items.forEach(section => {\n section.items.forEach(item => {\n // Extract impact level from the item if it exists\n let impact = 1; // Default low impact\n\n // Look for impact indicators in the item\n if (item.toLowerCase().includes(\"**high impact**\")) {\n impact = 3;\n } else if (item.toLowerCase().includes(\"**medium impact**\")) {\n impact = 2;\n } else {\n // No impact indicator\n const itemLower = item.toLowerCase();\n\n // Fallback if call fails\n if (\n itemLower.includes(\"major feature\")\n || itemLower.includes(\"significant improvement\")\n || itemLower.includes(\"important\")\n ) {\n impact = 3; // High impact\n } else if (\n itemLower.includes(\"improved\")\n || itemLower.includes(\"enhanced\")\n || itemLower.includes(\"new\")\n ) {\n impact = 2; // Medium impact\n }\n }\n\n allItems.push({ item, impact });\n });\n });\n\n // Sort by impact (high to low)\n allItems.sort((a, b) => b.impact - a.impact);\n\n // Rebuild the markdown with sorted items\n let sortedMarkdown = header + \"\\n\";\n sortedMarkdown += \"### Changes (Sorted by User Impact)\\n\";\n allItems.forEach(({ item }) => {\n // Remove the impact marker from the displayed text\n sortedMarkdown += cleanImpactIndicators(item) + \"\\n\";\n });\n\n return sortedMarkdown;\n }\n\n if (sortOption === \"chronological\") {\n // For chronological, we'll just reverse the items in each category\n let chronologicalMarkdown = header + \"\\n\";\n\n items.forEach(section => {\n chronologicalMarkdown += `### ${section.category}\\n`;\n\n // Reverse the items to show newest first\n const reversedItems = [...section.items].reverse();\n reversedItems.forEach(item => {\n // Clean impact indicators before adding to output\n chronologicalMarkdown += cleanImpactIndicators(item) + \"\\n\";\n });\n\n chronologicalMarkdown += \"\\n\";\n });\n\n return chronologicalMarkdown;\n }\n\n return markdown;\n } catch (error) {\n console.error(\"Error sorting markdown:\", error);\n return markdown;\n }\n };\n\n // Process PR references in the markdown\n const processPRRefs = (markdown: string): string => {\n if (!showPRLinks || !owner || !repo) {\n return markdown;\n }\n\n // Replace PR references with links\n // Look for patterns like \"PR #123\" or \"(#123)\" that aren't already part of a link\n return markdown.replace(\n /(? {\n return `[${match}](https://github.com/${owner}/${repo}/pull/${prNumber})`;\n },\n );\n };\n\n // Convert markdown to HTML\n const htmlContent = React.useMemo(() => {\n try {\n const processedMarkdown = processPRRefs(\n isEditing ? editableMarkdown : sortMarkdown(editableMarkdown, sortOption),\n );\n return { __html: marked(processedMarkdown) };\n } catch (error) {\n console.error(\"Error parsing markdown:\", error);\n return { __html: `

Error rendering markdown: ${error.message}

` };\n }\n }, [editableMarkdown, sortOption, showPRLinks, isEditing, owner, repo]);\n\n // Handle markdown changes\n const handleMarkdownChange = (value: string) => {\n setEditableMarkdown(value);\n if (onMarkdownChange) {\n onMarkdownChange(value);\n }\n };\n\n // Generate shareable link\n const generateShareableLink = async () => {\n if (!owner || !repo || !startDate || !endDate) {\n alert(\"Missing repository information. Please make sure you've selected a repository and date range.\");\n return;\n }\n\n setIsSharing(true);\n\n try {\n const response = await fetch(\"/api/share\", {\n method: \"POST\",\n headers: {\n \"Content-Type\": \"application/json\",\n },\n body: JSON.stringify({\n owner,\n repo,\n startDate,\n endDate,\n releaseNotes: editableMarkdown,\n }),\n });\n\n const data = await response.json();\n\n if (!response.ok) {\n throw new Error(data.error || \"Failed to create shareable link\");\n }\n\n // Create the shareable URL\n const shareUrl = new URL(window.location.href);\n shareUrl.search = `?share=${data.id}`;\n\n setShareableUrl(shareUrl.toString());\n setShowShareModal(true);\n } catch (error) {\n console.error(\"Error creating shareable link:\", error);\n alert(`Failed to create shareable link: ${error.message}`);\n } finally {\n setIsSharing(false);\n }\n };\n\n // Add custom CSS for markdown rendering\n const markdownStyles = `\n .markdown-preview h1 { font-size: 1.8rem; font-weight: bold; margin-top: 1.5rem; margin-bottom: 1rem; }\n .markdown-preview h2 { font-size: 1.5rem; font-weight: bold; margin-top: 1.5rem; margin-bottom: 1rem; }\n .markdown-preview h3 { font-size: 1.3rem; font-weight: bold; margin-top: 1.2rem; margin-bottom: 0.8rem; }\n .markdown-preview h4 { font-size: 1.1rem; font-weight: bold; margin-top: 1rem; margin-bottom: 0.6rem; }\n .markdown-preview p { margin-bottom: 1rem; }\n .markdown-preview ul { list-style-type: disc; margin-left: 1.5rem; margin-bottom: 1rem; }\n .markdown-preview ol { list-style-type: decimal; margin-left: 1.5rem; margin-bottom: 1rem; }\n .markdown-preview li { margin-bottom: 0.5rem; }\n .markdown-preview code { background-color: #f0f0f0; padding: 0.2rem 0.4rem; border-radius: 0.2rem; font-family: monospace; }\n .markdown-preview pre { background-color: #f0f0f0; padding: 1rem; border-radius: 0.4rem; overflow-x: auto; margin-bottom: 1rem; }\n .markdown-preview blockquote { border-left: 4px solid #e2e8f0; padding-left: 1rem; margin-left: 0; margin-right: 0; margin-bottom: 1rem; }\n .markdown-preview a { color: #3182ce; text-decoration: underline; }\n .markdown-preview table { border-collapse: collapse; width: 100%; margin-bottom: 1rem; }\n .markdown-preview th, .markdown-preview td { border: 1px solid #e2e8f0; padding: 0.5rem; }\n .markdown-preview th { background-color: #f7fafc; }\n .markdown-preview hr { border: 0; border-top: 1px solid #e2e8f0; margin: 1.5rem 0; }\n `;\n\n return (\n

\n \n\n

Generated Release Notes

\n setIsEditing(!isEditing)}\n className=\"text-sm bg-gray-200 px-3 py-1 rounded hover:bg-gray-300\"\n >\n {isEditing ? \"Preview\" : \"Edit\"}\n \n \n {isSharing ? \"Sharing...\" : \"Share\"}\n \n

\n\n {!isEditing && (\n

\n Sort by:\n setSortOption(e.target.value as SortOption)}\n className=\"p-1 border rounded text-sm\"\n >\n \n \n \n \n

\n\n

\n setShowPRLinks(e.target.checked)}\n className=\"mr-2\"\n />\n \n Link PR references\n \n

\n )}\n\n {isEditing\n ? (\n

\n handleMarkdownChange(e.target.value)}\n className=\"w-full h-96 font-mono text-sm p-2 border rounded focus:outline-none focus:ring-2 focus:ring-blue-500\"\n spellCheck=\"false\"\n />\n

\n )\n : (\n

\n \n

\n )}\n\n {/* Share Modal */}\n {showShareModal && (\n

Share these release notes

\n \n {\n navigator.clipboard.writeText(shareableUrl);\n setShareCopied(true);\n setTimeout(() => setShareCopied(false), 2000);\n }}\n className=\"bg-blue-500 text-white px-4 py-2 rounded-r\"\n >\n {shareCopied ? \"Copied!\" : \"Copy\"}\n \n

\n Anyone with this link can view these release notes.\n

\n setShowShareModal(false)}\n className=\"text-gray-600 hover:text-gray-800\"\n >\n Close\n \n

\n )}\n

\n );\n}",[],[147,228,230],"8743ce22-2c1c-11f0-a285-569c3dd06744","CommitList.tsx","8743ced6-2c1c-11f0-a285-569c3dd06744",["D",1746716148849],"/** @jsxImportSource https://esm.sh/react@18.2.0 */\nimport React, { useState } from \"https://esm.sh/react@18.2.0\";\nimport { CommitWithPR } from \"../../shared/types.ts\";\n\ninterface CommitListProps {\n commits: CommitWithPR[];\n}\n\nexport function CommitList({ commits }: CommitListProps) {\n const [expandedCommit, setExpandedCommit] = useState(null);\n\n const toggleExpand = (sha: string) => {\n if (expandedCommit === sha) {\n setExpandedCommit(null);\n } else {\n setExpandedCommit(sha);\n }\n };\n\n if (commits.length === 0) {\n return

No commits found in the selected date range.

;\n }\n\n return (\n

\n \n \n \n \n \n \n \n \n \n \n {commits.map((commit) => (\n \n toggleExpand(commit.sha)}\n >\n \n \n \n \n \n {expandedCommit === commit.sha && (\n \n \n \n )}\n \n ))}\n \n

Commit	Date	Author	PR
\n \n \n e.stopPropagation()}\n >\n {commit.sha.substring(0, 7)}\n \n \n \n {commit.message.split('\\n')[0]}\n \n \n	\n {new Date(commit.date).toLocaleDateString()}\n	\n \n {commit.authorUrl ? (\n e.stopPropagation()}\n >\n {commit.author}\n \n ) : (\n commit.author\n )}\n \n	\n {commit.pr ? (\n e.stopPropagation()}\n >\n #{commit.pr.number}\n \n ) : (\n \"-\"\n )}\n
\n \n Commit Message: \n \n {commit.message}\n \n \n {commit.pr && (\n <>\n PR Title: \n \n {commit.pr.title}\n \n \n {commit.pr.description && (\n <>\n PR Description: \n \n {commit.pr.description}\n \n \n )}\n \n )}\n \n

\n );\n}",[],[147,228,239],"7c0f42e8-2c1c-11f0-99b5-569c3dd06744",["D",1746716130048],[174,175,176],[147,228],["D",1746716093465],[173,177,178],[147],"8cf69880-2cdb-11f0-8ca1-569c3dd06744","llm.ts","623b7fbc-2c1c-11f0-aa08-569c3dd06744",51,["D",1746798192283],"import { OpenAI } from \"https://esm.town/v/std/openai\";\nimport { CommitWithPR, ReleaseNote } from \"../shared/types.ts\";\n\n// Initialize OpenAI client\nconst openai = new OpenAI();\n\n/**\n * Generates user-facing release notes from commit and PR data\n */\nexport async function generateReleaseNotes(commits: CommitWithPR[]): Promise<{\n notes: ReleaseNote[];\n markdown: string;\n}> {\n // Prepare data for LLM call\n const commitData = commits.map(commit => {\n return {\n sha: commit.sha.substring(0, 7),\n message: commit.message,\n date: commit.date,\n author: commit.author,\n pr: commit.pr\n ? {\n number: commit.pr.number,\n title: commit.pr.title,\n description: commit.pr.description || \"\",\n }\n : null,\n };\n });\n\n // Chunk data\n const MAX_COMMITS_PER_CHUNK = 50;\n const commitChunks = [];\n\n for (let i = 0; i < commitData.length; i += MAX_COMMITS_PER_CHUNK) {\n commitChunks.push(commitData.slice(i, i + MAX_COMMITS_PER_CHUNK));\n }\n\n console.log(`Processing ${commitData.length} commits in ${commitChunks.length} chunks`);\n\n // Process each chunk and combine the results\n const chunkResults = [];\n\n for (let i = 0; i < commitChunks.length; i++) {\n const chunk = commitChunks[i];\n console.log(`Processing chunk ${i + 1}/${commitChunks.length} with ${chunk.length} commits`);\n\n const chunkResult = await processCommitChunk(chunk, i, commitChunks.length);\n chunkResults.push(chunkResult);\n }\n\n // Combine the results from all chunks\n let allNotes: ReleaseNote[] = [];\n let combinedMarkdown = \"\";\n\n for (const result of chunkResults) {\n allNotes = [...allNotes, ...result.notes];\n\n if (combinedMarkdown && result.markdown) {\n // Extract sections from the markdown to avoid duplicating headers\n const sectionRegex = /### (Features|Improvements|Bug Fixes|Other Changes)([\\s\\S]*?)(?=### |$)/g;\n let match;\n\n while ((match = sectionRegex.exec(result.markdown)) !== null) {\n const sectionType = match[1];\n const sectionContent = match[2].trim();\n\n if (sectionContent) {\n // Check if this section already exists in the combined markdown\n const sectionHeader = `### ${sectionType}`;\n if (combinedMarkdown.includes(sectionHeader)) {\n // Append to existing section\n combinedMarkdown = combinedMarkdown.replace(\n new RegExp(`(### ${sectionType}[\\\\\\\\s\\\\\\\\S]*?)(?=### |$)`),\n `$1\\n${sectionContent}\\n\\n`,\n );\n } else {\n // Add new section\n combinedMarkdown += `\\n\\n### ${sectionType}\\n${sectionContent}\\n`;\n }\n }\n }\n } else if (result.markdown) {\n combinedMarkdown = result.markdown;\n }\n }\n\n // If we have multiple chunks, we might need to deduplicate and reorganize\n if (commitChunks.length > 1) {\n // Deduplicate notes by title\n const uniqueNotes: Record = {};\n for (const note of allNotes) {\n const key = note.title.toLowerCase();\n if (!uniqueNotes[key] || note.originalCommits.length > uniqueNotes[key].originalCommits.length) {\n uniqueNotes[key] = note;\n }\n }\n allNotes = Object.values(uniqueNotes);\n\n // If we have deduped notes, regenerate the markdown\n if (allNotes.length < chunkResults.reduce((sum, r) => sum + r.notes.length, 0)) {\n combinedMarkdown = generateMarkdownFromNotes(allNotes);\n }\n }\n\n return {\n notes: allNotes,\n markdown: combinedMarkdown,\n };\n}\n\n/**\n * Process a chunk of commits to generate release notes\n */\nasync function processCommitChunk(\n commitChunk: any[],\n chunkIndex: number,\n totalChunks: number,\n): Promise<{\n notes: ReleaseNote[];\n markdown: string;\n}> {\n // Create the prompt for the LLM\n const prompt = `\nYou are a technical writer creating release notes for a software product. \nI'll provide you with a list of commits and pull requests. \nYour task is to generate clear, concise, developer-friendly release notes.\n\n${totalChunks > 1 ? `This is chunk ${chunkIndex + 1} of ${totalChunks} from the full commit history.` : \"\"}\n${totalChunks > 1 ? \"Make sure to consider ALL commits in this chunk, not just the most recent ones.\" : \"\"}\n\nGuidelines:\n1. Focus ONLY on changes that are relevant to end-users or developers\n2. Ignore internal refactoring, technical debt, or backend changes that don't affect user experience\n3. Ignore work that isn't yet released or incomplete features\n4. Group similar changes together\n5. Use clear, developer-friendly language (like PostHog's style - concise, technical but accessible)\n6. Be concise but descriptive\n7. IMPORTANT: Consider ALL commits in the provided list, not just recent ones\n8. When a change is associated with a PR, include \"PR #[number]\" in the description\n\nFor impact assessment, use these specific criteria:\n- \"high\" impact: Major new features that users will love and significantly improve their experience\n- \"medium\" impact: Nice-to-have new features and substantial improvements to existing functionality\n- \"low\" impact: Minor improvements and bug fixes that users will notice\n- DO NOT include changes that aren't user-facing, heavy backend changes, or very small fixes\n\nFor each relevant change, include:\n- A clear title describing the change\n- A brief description explaining the benefit to users/developers\n- Categorize as: \"Feature\", \"Improvement\", \"Bugfix\", or \"Other\"\n- Assign an impact level based on the criteria above\n- Include PR references in the description when available\n\nFormat your response in two parts:\n1. First, a JSON array of release notes with the structure:\n\\`\\`\\`json\n[\n {\n \"title\": \"Added new dashboard widgets\",\n \"description\": \"New visualization options for better data insights (PR #123)\",\n \"type\": \"feature\",\n \"impact\": \"high\",\n \"originalCommits\": [\"abc1234\", \"def5678\"]\n },\n ...\n]\n\\`\\`\\`\n\n2. Second, a markdown formatted version of the release notes:\n\\`\\`\\`markdown\n## Release Notes\n\n### Features\n- **Added new dashboard widgets**: New visualization options for better data insights (PR #123)\n\n### Improvements\n- **Faster page loading**: Optimized image loading for better performance (PR #456)\n\n### Bug Fixes\n- **Fixed login issue**: Resolved problem with social login providers (PR #789)\n\\`\\`\\`\n\nHere are the commits and PRs to analyze:\n${JSON.stringify(commitChunk, null, 2)}\n`;\n\n // Call the OpenAI API\n const completion = await openai.chat.completions.create({\n model: \"gpt-4o\",\n messages: [\n { role: \"system\", content: \"You are a technical writer creating comprehensive release notes for developers.\" },\n { role: \"user\", content: prompt },\n ],\n temperature: 0.3,\n max_tokens: 4000, // Increased token limit for larger responses\n });\n\n const response = completion.choices[0].message.content || \"\";\n\n // Extract the JSON and markdown parts\n let notes: ReleaseNote[] = [];\n let markdown = \"\";\n\n try {\n // Extract JSON\n const jsonMatch = response.match(/\\`\\`\\`json\\n([\\s\\S]*?)\\n\\`\\`\\`/);\n if (jsonMatch && jsonMatch[1]) {\n notes = JSON.parse(jsonMatch[1]);\n } else {\n console.warn(\"Could not extract JSON from LLM response\");\n }\n\n // Extract markdown\n const markdownMatch = response.match(/\\`\\`\\`markdown\\n([\\s\\S]*?)\\n\\`\\`\\`/);\n if (markdownMatch && markdownMatch[1]) {\n markdown = markdownMatch[1];\n } else {\n console.warn(\"Could not extract markdown from LLM response\");\n if (notes.length > 0) {\n markdown = generateMarkdownFromNotes(notes);\n }\n }\n } catch (error) {\n console.error(\"Error parsing LLM response:\", error);\n throw new Error(\"Failed to parse LLM response: \" + error.message);\n }\n\n return {\n notes,\n markdown,\n };\n}\n\n/**\n * Generate markdown from release notes\n */\nfunction generateMarkdownFromNotes(notes: ReleaseNote[]): string {\n // Group notes by type\n const groupedNotes: Record = {\n feature: [],\n improvement: [],\n bugfix: [],\n other: [],\n };\n\n for (const note of notes) {\n const type = note.type.toLowerCase() as keyof typeof groupedNotes;\n if (groupedNotes[type]) {\n groupedNotes[type].push(note);\n } else {\n groupedNotes.other.push(note);\n }\n }\n\n // Generate markdown\n let markdown = \"## Release Notes\\n\\n\";\n\n // Features\n if (groupedNotes.feature.length > 0) {\n markdown += \"### Features\\n\";\n for (const note of groupedNotes.feature) {\n // Include impact level as a hidden marker for frontend sorting\n const impactMarker = note.impact ? `**${note.impact} impact**` : \"\";\n markdown += `- **${note.title}**: ${note.description} ${impactMarker}\\n`;\n }\n markdown += \"\\n\";\n }\n\n // Improvements\n if (groupedNotes.improvement.length > 0) {\n markdown += \"### Improvements\\n\";\n for (const note of groupedNotes.improvement) {\n const impactMarker = note.impact ? `**${note.impact} impact**` : \"\";\n markdown += `- **${note.title}**: ${note.description} ${impactMarker}\\n`;\n }\n markdown += \"\\n\";\n }\n\n // Bug Fixes\n if (groupedNotes.bugfix.length > 0) {\n markdown += \"### Bug Fixes\\n\";\n for (const note of groupedNotes.bugfix) {\n const impactMarker = note.impact ? `**${note.impact} impact**` : \"\";\n markdown += `- **${note.title}**: ${note.description} ${impactMarker}\\n`;\n }\n markdown += \"\\n\";\n }\n\n // Other Changes\n if (groupedNotes.other.length > 0) {\n markdown += \"### Other Changes\\n\";\n for (const note of groupedNotes.other) {\n const impactMarker = note.impact ? `**${note.impact} impact**` : \"\";\n markdown += `- **${note.title}**: ${note.description} ${impactMarker}\\n`;\n }\n markdown += \"\\n\";\n }\n\n return markdown.trim();\n}",[],[152,253],"0c443678-2cdd-11f0-982f-569c3dd06744","github.ts","58043b9c-2c1c-11f0-b7ff-569c3dd06744",57,["D",1746798835358],14,"import { CommitWithPR, GitHubCommit, GitHubPR } from \"../shared/types.ts\";\n\n/**\n * Fetches commits from a GitHub repository within a date range\n * Implements pagination to fetch all commits\n */\nexport async function fetchCommits(\n token: string,\n owner: string,\n repo: string,\n since: string,\n until: string,\n): Promise {\n // Format dates as ISO 8601 if they're not already\n const sinceDate = new Date(since);\n const untilDate = new Date(until);\n\n // Add time component to make it a full ISO string and handle timezone\n const sinceISO = sinceDate.toISOString();\n // Set until date to end of day\n untilDate.setHours(23, 59, 59, 999);\n const untilISO = untilDate.toISOString();\n\n let allCommits: GitHubCommit[] = [];\n let page = 1;\n let hasMorePages = true;\n\n // GitHub API headers\n const headers = {\n \"Accept\": \"application/vnd.github.v3+json\",\n \"Authorization\": `Bearer ${token}`,\n \"User-Agent\": \"GitHub-Release-Notes-Generator\",\n \"X-GitHub-Api-Version\": \"2022-11-28\",\n };\n\n console.log(`Fetching commits between ${sinceISO} and ${untilISO}`);\n\n // Fetch all pages\n while (hasMorePages) {\n const url =\n `https://api.github.com/repos/${owner}/${repo}/commits?since=${sinceISO}&until=${untilISO}&per_page=100&page=${page}`;\n console.log(`Fetching page ${page}...`);\n\n const response = await fetch(url, { headers });\n\n if (!response.ok) {\n const error = await response.text();\n throw new Error(`GitHub API error: ${response.status} - ${error}`);\n }\n\n const commits = await response.json();\n\n if (commits.length === 0) {\n hasMorePages = false;\n } else {\n allCommits = [...allCommits, ...commits];\n\n // Check if there's a next page using the Link header\n const linkHeader = response.headers.get(\"Link\");\n hasMorePages = linkHeader ? linkHeader.includes(\"rel=\\\"next\\\"\") : false;\n\n page++;\n }\n }\n\n console.log(`Fetched a total of ${allCommits.length} commits`);\n return allCommits;\n}\n\n/**\n * Fetches a pull request by its number\n */\nexport async function fetchPR(\n token: string,\n owner: string,\n repo: string,\n prNumber: number,\n): Promise {\n const url = `https://api.github.com/repos/${owner}/${repo}/pulls/${prNumber}`;\n\n const response = await fetch(url, {\n headers: {\n \"Accept\": \"application/vnd.github.v3+json\",\n \"Authorization\": `Bearer ${token}`,\n \"User-Agent\": \"GitHub-Release-Notes-Generator\",\n \"X-GitHub-Api-Version\": \"2022-11-28\",\n },\n });\n\n if (response.status === 404) {\n return null;\n }\n\n if (!response.ok) {\n const error = await response.text();\n throw new Error(`GitHub API error: ${response.status} - ${error}`);\n }\n\n return response.json();\n}\n\n/**\n * Fetches pull requests associated with a commit.\n */\nasync function fetchPRsForCommit(\n token: string,\n owner: string,\n repo: string,\n commitSha: string,\n): Promise {\n const url = `https://api.github.com/repos/${owner}/${repo}/commits/${commitSha}/pulls`;\n\n const response = await fetch(url, {\n headers: {\n \"Accept\": \"application/vnd.github.v3+json\",\n \"Authorization\": `Bearer ${token}`,\n \"User-Agent\": \"GitHub-Release-Notes-Generator\",\n \"X-GitHub-Api-Version\": \"2022-11-28\",\n },\n });\n\n if (response.status === 404) {\n return [];\n }\n\n if (!response.ok) {\n const error = await response.text();\n throw new Error(`GitHub API error: ${response.status} - ${error}`);\n }\n\n return response.json();\n}\n\n/**\n * Fetches issues referenced by a commit or PR using GitHub API\n */\nasync function fetchIssueReferences(\n token: string,\n owner: string,\n repo: string,\n commitSha: string,\n prNumber?: number,\n): Promise {\n // First, try to get issues that are closed by this commit or PR\n const issueNumbers: number[] = [];\n\n // If we have a PR number, check for issues closed by the PR\n if (prNumber) {\n const prIssuesUrl = `https://api.github.com/repos/${owner}/${repo}/pulls/${prNumber}/commits`;\n\n const prResponse = await fetch(prIssuesUrl, {\n headers: {\n \"Accept\": \"application/vnd.github.v3+json\",\n \"Authorization\": `Bearer ${token}`,\n \"User-Agent\": \"GitHub-Release-Notes-Generator\",\n \"X-GitHub-Api-Version\": \"2022-11-28\",\n },\n });\n\n if (prResponse.ok) {\n const prCommits = await prResponse.json();\n\n // Extract issue numbers from commit messages using the GitHub API pattern\n for (const commit of prCommits) {\n // Look for \"Fixes #X\", \"Closes #X\", \"Resolves #X\" patterns that GitHub recognizes\n const message = commit.commit.message;\n const keywords = [\"close\", \"closes\", \"closed\", \"fix\", \"fixes\", \"fixed\", \"resolve\", \"resolves\", \"resolved\"];\n\n for (const keyword of keywords) {\n const regex = new RegExp(`${keyword}\\\\s+#(\\\\d+)`, \"gi\");\n let match;\n while ((match = regex.exec(message)) !== null) {\n issueNumbers.push(parseInt(match[1], 10));\n }\n }\n }\n }\n }\n\n // Also check the timeline of the PR to find linked issues\n if (prNumber) {\n const timelineUrl = `https://api.github.com/repos/${owner}/${repo}/issues/${prNumber}/timeline`;\n\n const timelineResponse = await fetch(timelineUrl, {\n headers: {\n \"Accept\": \"application/vnd.github.mockingbird-preview+json\",\n \"Authorization\": `Bearer ${token}`,\n \"User-Agent\": \"GitHub-Release-Notes-Generator\",\n \"X-GitHub-Api-Version\": \"2022-11-28\",\n },\n });\n\n if (timelineResponse.ok) {\n const timeline = await timelineResponse.json();\n\n // Look for cross-references and connected events\n for (const event of timeline) {\n if (event.event === \"cross-referenced\" && event.source?.issue?.number) {\n issueNumbers.push(event.source.issue.number);\n }\n if (event.event === \"connected\" && event.issue?.number) {\n issueNumbers.push(event.issue.number);\n }\n }\n }\n }\n\n // Remove duplicates\n return [...new Set(issueNumbers)];\n}\n\n/**\n * Fetches commits and their associated PRs\n * Includes progress tracking for large repositories\n * Uses GitHub API to find PR and issue references\n */\nexport async function fetchCommitsWithPRs(\n token: string,\n owner: string,\n repo: string,\n startDate: string,\n endDate: string,\n): Promise {\n // Fetch all commits with pagination\n const commits = await fetchCommits(token, owner, repo, startDate, endDate);\n\n const commitsWithPRs: CommitWithPR[] = [];\n const totalCommits = commits.length;\n\n console.log(`Processing ${totalCommits} commits to extract PR information...`);\n\n // Process commits in batches to not get rate limited\n const BATCH_SIZE = 5; // Reduced batch size since we're making more API calls per commit\n for (let i = 0; i < commits.length; i += BATCH_SIZE) {\n const batch = commits.slice(i, i + BATCH_SIZE);\n const batchPromises = batch.map(async (commit) => {\n // Use GitHub API to find associated PRs for this commit\n const associatedPRs = await fetchPRsForCommit(token, owner, repo, commit.sha);\n\n let pr: GitHubPR | null = null;\n let issueRefs: number[] = [];\n\n // If we found associated PRs, use the first one\n if (associatedPRs.length > 0) {\n pr = associatedPRs[0]; // Use the first PR if multiple exist\n\n // Get issue references using the GitHub API\n const prIssueRefs = await fetchIssueReferences(token, owner, repo, commit.sha, pr.number);\n issueRefs = [...prIssueRefs];\n } else {\n // If no PR is associated, still check for issue references in the commit\n const commitIssueRefs = await fetchIssueReferences(token, owner, repo, commit.sha);\n issueRefs = [...commitIssueRefs];\n }\n\n return {\n sha: commit.sha,\n message: commit.commit.message,\n date: commit.commit.author.date,\n author: commit.author?.login || commit.commit.author.name,\n authorUrl: commit.author?.html_url || \"\",\n commitUrl: commit.html_url,\n issueRefs: issueRefs.length > 0 ? issueRefs : undefined,\n ...(pr && {\n pr: {\n number: pr.number,\n title: pr.title,\n description: pr.body,\n url: pr.html_url,\n },\n }),\n };\n });\n\n const batchResults = await Promise.all(batchPromises);\n commitsWithPRs.push(...batchResults);\n\n console.log(\n `Processed ${Math.min(i + BATCH_SIZE, totalCommits)}/${totalCommits} commits (${\n Math.round((Math.min(i + BATCH_SIZE, totalCommits) / totalCommits) * 100)\n }%)`,\n );\n }\n\n // Sort commits by date (newest first)\n commitsWithPRs.sort((a, b) => new Date(b.date).getTime() - new Date(a.date).getTime());\n\n return commitsWithPRs;\n}",[],[152,261],"9b3ea140-2c1c-11f0-800c-569c3dd06744","9b3ea1fe-2c1c-11f0-800c-569c3dd06744",13,["D",1746716182368],"# Backend\n\nThis directory contains the server-side code for the GitHub Release Notes Generator.\n\n## Files\n\n- `github.ts` - GitHub API integration for fetching commits and PR data\n- `llm.ts` - OpenAI integration for generating release notes\n- `types.ts` - Type definitions for the backend\n\n## API Endpoints\n\n### POST /api/commits\n\nFetches commits and associated PR data from GitHub within a specified date range.\n\n**Request Body:**\n```json\n{\n \"token\": \"github_personal_access_token\",\n \"owner\": \"repository_owner\",\n \"repo\": \"repository_name\",\n \"startDate\": \"2023-01-01\",\n \"endDate\": \"2023-01-31\"\n}\n```\n\n**Response:**\n```json\n{\n \"commits\": [\n {\n \"sha\": \"commit_sha\",\n \"message\": \"commit_message\",\n \"date\": \"commit_date\",\n \"author\": \"author_name\",\n \"authorUrl\": \"author_github_url\",\n \"commitUrl\": \"commit_url\",\n \"pr\": {\n \"number\": 123,\n \"title\": \"PR title\",\n \"description\": \"PR description\",\n \"url\": \"PR URL\"\n }\n }\n ]\n}\n```\n\n### POST /api/generate-notes\n\nGenerates user-facing release notes from commit and PR data using OpenAI.\n\n**Request Body:**\n```json\n{\n \"commits\": [\n // Array of commit objects from /api/commits\n ]\n}\n```\n\n**Response:**\n```json\n{\n \"releaseNotes\": {\n \"notes\": [\n {\n \"title\": \"Feature title\",\n \"description\": \"Feature description\",\n \"type\": \"feature\",\n \"originalCommits\": [\"commit_sha1\", \"commit_sha2\"]\n }\n ],\n \"markdown\": \"## Release Notes\\n\\n### Features\\n- **Feature title**: Feature description\"\n }\n}\n```",[],[152,132],["D",1746716069578],[169,170,171],[152],["D",1746715516569],{"_16":17,"_85":81},{"_16":17,"_79":-5,"_80":81,"_34":35,"_82":83,"_84":26},{"_67":68,"_20":69,"_18":70,"_71":19,"_72":-5,"_73":-5,"_74":-5,"_75":-5,"_76":-5,"_77":21,"_78":17},"ctx",{"_60":-5,"_106":316,"_317":318,"_319":320,"_321":26,"_322":26,"_323":26,"_324":39,"_325":-5,"_326":-7,"_327":-7,"_328":44,"_329":26,"_330":26,"_331":-5,"_332":26,"_333":26,"_334":26,"_335":-5},"ENV",{"_287":288,"_289":290,"_291":69,"_292":293,"_294":295,"_296":297,"_298":299,"_300":301,"_302":303,"_304":305,"_306":307,"_308":309,"_310":311,"_312":313,"_314":315},"REACT_APP_SENTRY_DSN","https://586020d86f05489db63a554690acf1e9@o1398359.ingest.sentry.io/6724499","RENDER_GIT_COMMIT","5015b8df46402d8bba4f485ecc37aec202b0e06b","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"]

Search