#!/usr/bin/env node
import fs from "fs";
import path, { dirname } from "path";
import { fileURLToPath } from "url";
import {
AGENTS_DIR,
AKA_INSTALL_URLS,
COLLECTIONS_DIR,
DOCS_DIR,
INSTRUCTIONS_DIR,
PROMPTS_DIR,
repoBaseUrl,
ROOT_FOLDER,
SKILLS_DIR,
TEMPLATES,
vscodeInsidersInstallImage,
vscodeInstallImage,
} from "./constants.mjs";
import {
extractMcpServerConfigs,
parseCollectionYaml,
parseFrontmatter,
parseSkillMetadata,
} from "./yaml-parser.mjs";
const __filename = fileURLToPath(import.meta.url);
const __dirname = dirname(__filename);
// Cache of MCP registry server names (lower-cased) fetched from the API
let MCP_REGISTRY_SET = null;
/**
* Loads and caches the set of MCP registry server names from the GitHub MCP registry API.
*
* Behavior:
* - If a cached set already exists (MCP_REGISTRY_SET), it is returned immediately.
* - Fetches all pages from https://api.mcp.github.com/v0.1/servers/ using cursor-based pagination
* - Safely handles network errors or malformed JSON by returning an empty array.
* - Extracts server names from: data[].server.name
* - Normalizes names to lowercase for case-insensitive matching
* - Only hits the API once per README build run (cached for subsequent calls)
*
* Side Effects:
* - Mutates the module-scoped variable MCP_REGISTRY_SET.
* - Logs a warning to console if fetching or parsing the registry fails.
*
* @returns {Promise<{ name: string, displayName: string }[]>} Array of server entries with name and lowercase displayName. May be empty if
* the API is unreachable or returns malformed data.
*
* @throws {none} All errors are caught internally; failures result in an empty array.
*/
async function loadMcpRegistryNames() {
if (MCP_REGISTRY_SET) return MCP_REGISTRY_SET;
try {
console.log("Fetching MCP registry from API...");
const allServers = [];
let cursor = null;
const apiUrl = "https://api.mcp.github.com/v0.1/servers/";
// Fetch all pages using cursor-based pagination
do {
const url = cursor
? `${apiUrl}?cursor=${encodeURIComponent(cursor)}`
: apiUrl;
const response = await fetch(url);
if (!response.ok) {
throw new Error(`API returned status ${response.status}`);
}
const json = await response.json();
const servers = json?.servers || [];
// Extract server names and displayNames from the response
for (const entry of servers) {
const serverName = entry?.server?.name;
if (serverName) {
// Try to get displayName from GitHub metadata, fall back to server name
const displayName =
entry?.server?._meta?.[
"io.modelcontextprotocol.registry/publisher-provided"
]?.github?.displayName || serverName;
allServers.push({
name: serverName,
displayName: displayName.toLowerCase(),
// Also store the original full name for matching
fullName: serverName.toLowerCase(),
});
}
}
// Get next cursor for pagination
cursor = json?.metadata?.nextCursor || null;
} while (cursor);
console.log(`Loaded ${allServers.length} servers from MCP registry`);
MCP_REGISTRY_SET = allServers;
} catch (e) {
console.warn(`Failed to load MCP registry from API: ${e.message}`);
MCP_REGISTRY_SET = [];
}
return MCP_REGISTRY_SET;
}
// Add error handling utility
/**
* Safe file operation wrapper
*/
function safeFileOperation(operation, filePath, defaultValue = null) {
try {
return operation();
} catch (error) {
console.error(`Error processing file ${filePath}: ${error.message}`);
return defaultValue;
}
}
function extractTitle(filePath) {
return safeFileOperation(
() => {
const content = fs.readFileSync(filePath, "utf8");
const lines = content.split("\n");
// Step 1: Try to get title from frontmatter using vfile-matter
const frontmatter = parseFrontmatter(filePath);
if (frontmatter) {
// Check for title field
if (frontmatter.title && typeof frontmatter.title === "string") {
return frontmatter.title;
}
// Check for name field and convert to title case
if (frontmatter.name && typeof frontmatter.name === "string") {
return frontmatter.name
.split("-")
.map((word) => word.charAt(0).toUpperCase() + word.slice(1))
.join(" ");
}
}
// Step 2: For prompt/agent/instructions files, look for heading after frontmatter
if (
filePath.includes(".prompt.md") ||
filePath.includes(".agent.md") ||
filePath.includes(".instructions.md")
) {
// Look for first heading after frontmatter
let inFrontmatter = false;
let frontmatterEnded = false;
let inCodeBlock = false;
for (const line of lines) {
if (line.trim() === "---") {
if (!inFrontmatter) {
inFrontmatter = true;
} else if (inFrontmatter && !frontmatterEnded) {
frontmatterEnded = true;
}
continue;
}
// Only look for headings after frontmatter ends
if (frontmatterEnded || !inFrontmatter) {
// Track code blocks to ignore headings inside them
if (
line.trim().startsWith("```") ||
line.trim().startsWith("````")
) {
inCodeBlock = !inCodeBlock;
continue;
}
if (!inCodeBlock && line.startsWith("# ")) {
return line.substring(2).trim();
}
}
}
// Step 3: Format filename for prompt/chatmode/instructions files if no heading found
const basename = path.basename(
filePath,
filePath.includes(".prompt.md")
? ".prompt.md"
: filePath.includes(".agent.md")
? ".agent.md"
: ".instructions.md"
);
return basename
.replace(/[-_]/g, " ")
.replace(/\b\w/g, (l) => l.toUpperCase());
}
// Step 4: For other files, look for the first heading (but not in code blocks)
let inCodeBlock = false;
for (const line of lines) {
if (line.trim().startsWith("```") || line.trim().startsWith("````")) {
inCodeBlock = !inCodeBlock;
continue;
}
if (!inCodeBlock && line.startsWith("# ")) {
return line.substring(2).trim();
}
}
// Step 5: Fallback to filename
const basename = path.basename(filePath, path.extname(filePath));
return basename
.replace(/[-_]/g, " ")
.replace(/\b\w/g, (l) => l.toUpperCase());
},
filePath,
path
.basename(filePath, path.extname(filePath))
.replace(/[-_]/g, " ")
.replace(/\b\w/g, (l) => l.toUpperCase())
);
}
function extractDescription(filePath) {
return safeFileOperation(
() => {
// Use vfile-matter to parse frontmatter for all file types
const frontmatter = parseFrontmatter(filePath);
if (frontmatter && frontmatter.description) {
return frontmatter.description;
}
return null;
},
filePath,
null
);
}
/**
* Format arbitrary multiline text for safe rendering inside a markdown table cell.
* - Preserves line breaks by converting to
* - Escapes pipe characters (|) to avoid breaking table columns
* - Trims leading/trailing whitespace on each line
* - Collapses multiple consecutive blank lines
* This should be applied to descriptions across all file types when used in tables.
*
* @param {string|null|undefined} text
* @returns {string} table-safe content
*/
function formatTableCell(text) {
if (text === null || text === undefined) return "";
let s = String(text);
// Normalize line endings
s = s.replace(/\r\n/g, "\n");
// Split lines, trim, drop empty groups while preserving intentional breaks
const lines = s
.split("\n")
.map((l) => l.trim())
.filter((_, idx, arr) => {
// Keep single blank lines, drop consecutive blanks
if (arr[idx] !== "") return true;
return arr[idx - 1] !== ""; // allow one blank, remove duplicates
});
s = lines.join("\n");
// Escape table pipes
s = s.replace(/\|/g, "|");
// Convert remaining newlines to
for a single-cell rendering
s = s.replace(/\n/g, "
");
return s.trim();
}
function makeBadges(link, type) {
const aka = AKA_INSTALL_URLS[type] || AKA_INSTALL_URLS.instructions;
const vscodeUrl = `${aka}?url=${encodeURIComponent(
`vscode:chat-${type}/install?url=${repoBaseUrl}/${link}`
)}`;
const insidersUrl = `${aka}?url=${encodeURIComponent(
`vscode-insiders:chat-${type}/install?url=${repoBaseUrl}/${link}`
)}`;
return `[](${vscodeUrl})
[](${insidersUrl})`;
}
/**
* Generate the instructions section with a table of all instructions
*/
function generateInstructionsSection(instructionsDir) {
// Check if directory exists
if (!fs.existsSync(instructionsDir)) {
return "";
}
// Get all instruction files
const instructionFiles = fs
.readdirSync(instructionsDir)
.filter((file) => file.endsWith(".instructions.md"));
// Map instruction files to objects with title for sorting
const instructionEntries = instructionFiles.map((file) => {
const filePath = path.join(instructionsDir, file);
const title = extractTitle(filePath);
return { file, filePath, title };
});
// Sort by title alphabetically
instructionEntries.sort((a, b) => a.title.localeCompare(b.title));
console.log(`Found ${instructionEntries.length} instruction files`);
// Return empty string if no files found
if (instructionEntries.length === 0) {
return "";
}
// Create table header
let instructionsContent =
"| Title | Description |\n| ----- | ----------- |\n";
// Generate table rows for each instruction file
for (const entry of instructionEntries) {
const { file, filePath, title } = entry;
const link = encodeURI(`instructions/${file}`);
// Check if there's a description in the frontmatter
const customDescription = extractDescription(filePath);
// Create badges for installation links
const badges = makeBadges(link, "instructions");
if (customDescription && customDescription !== "null") {
// Use the description from frontmatter, table-safe
instructionsContent += `| [${title}](../${link})
${badges} | ${formatTableCell(
customDescription
)} |\n`;
} else {
// Fallback to the default approach - use last word of title for description, removing trailing 's' if present
const topic = title.split(" ").pop().replace(/s$/, "");
instructionsContent += `| [${title}](../${link})
${badges} | ${topic} specific coding standards and best practices |\n`;
}
}
return `${TEMPLATES.instructionsSection}\n${TEMPLATES.instructionsUsage}\n\n${instructionsContent}`;
}
/**
* Generate the prompts section with a table of all prompts
*/
function generatePromptsSection(promptsDir) {
// Check if directory exists
if (!fs.existsSync(promptsDir)) {
return "";
}
// Get all prompt files
const promptFiles = fs
.readdirSync(promptsDir)
.filter((file) => file.endsWith(".prompt.md"));
// Map prompt files to objects with title for sorting
const promptEntries = promptFiles.map((file) => {
const filePath = path.join(promptsDir, file);
const title = extractTitle(filePath);
return { file, filePath, title };
});
// Sort by title alphabetically
promptEntries.sort((a, b) => a.title.localeCompare(b.title));
console.log(`Found ${promptEntries.length} prompt files`);
// Return empty string if no files found
if (promptEntries.length === 0) {
return "";
}
// Create table header
let promptsContent = "| Title | Description |\n| ----- | ----------- |\n";
// Generate table rows for each prompt file
for (const entry of promptEntries) {
const { file, filePath, title } = entry;
const link = encodeURI(`prompts/${file}`);
// Check if there's a description in the frontmatter
const customDescription = extractDescription(filePath);
// Create badges for installation links
const badges = makeBadges(link, "prompt");
if (customDescription && customDescription !== "null") {
promptsContent += `| [${title}](../${link})
${badges} | ${formatTableCell(
customDescription
)} |\n`;
} else {
promptsContent += `| [${title}](../${link})
${badges} | | |\n`;
}
}
return `${TEMPLATES.promptsSection}\n${TEMPLATES.promptsUsage}\n\n${promptsContent}`;
}
/**
* Generate MCP server links for an agent
* @param {string[]} servers - Array of MCP server names
* @param {{ name: string, displayName: string }[]} registryNames - Pre-loaded registry names to avoid async calls
* @returns {string} - Formatted MCP server links with badges
*/
function generateMcpServerLinks(servers, registryNames) {
if (!servers || servers.length === 0) {
return "";
}
const badges = [
{
type: "vscode",
url: "https://img.shields.io/badge/Install-VS_Code-0098FF?style=flat-square",
badgeUrl: (serverName) =>
`https://aka.ms/awesome-copilot/install/mcp-vscode?vscode:mcp/by-name/${serverName}/mcp-server`,
},
{
type: "insiders",
url: "https://img.shields.io/badge/Install-VS_Code_Insiders-24bfa5?style=flat-square",
badgeUrl: (serverName) =>
`https://aka.ms/awesome-copilot/install/mcp-vscode?vscode-insiders:mcp/by-name/${serverName}/mcp-server`,
},
{
type: "visualstudio",
url: "https://img.shields.io/badge/Install-Visual_Studio-C16FDE?style=flat-square",
badgeUrl: (serverName) =>
`https://aka.ms/awesome-copilot/install/mcp-visualstudio?vscode:mcp/by-name/${serverName}/mcp-server`,
},
];
return servers
.map((entry) => {
// Support either a string name or an object with config
const serverObj = typeof entry === "string" ? { name: entry } : entry;
const serverName = String(serverObj.name).trim();
// Build config-only JSON (no name/type for stdio; just command+args+env)
let configPayload = {};
if (serverObj.type && serverObj.type.toLowerCase() === "http") {
// HTTP: url + headers
configPayload = {
url: serverObj.url || "",
headers: serverObj.headers || {},
};
} else {
// Local/stdio: command + args + env
configPayload = {
command: serverObj.command || "",
args: Array.isArray(serverObj.args)
? serverObj.args.map(encodeURIComponent)
: [],
env: serverObj.env || {},
};
}
const encodedConfig = encodeURIComponent(JSON.stringify(configPayload));
const installBadgeUrls = [
`[](https://aka.ms/awesome-copilot/install/mcp-vscode?name=${serverName}&config=${encodedConfig})`,
`[](https://aka.ms/awesome-copilot/install/mcp-vscodeinsiders?name=${serverName}&config=${encodedConfig})`,
`[](https://aka.ms/awesome-copilot/install/mcp-visualstudio/mcp-install?${encodedConfig})`,
].join("
");
// Match against both displayName and full name (case-insensitive)
const serverNameLower = serverName.toLowerCase();
const registryEntry = registryNames.find((entry) => {
// Exact match on displayName or fullName
if (
entry.displayName === serverNameLower ||
entry.fullName === serverNameLower
) {
return true;
}
// Check if the serverName matches a part of the full name after a slash
// e.g., "apify" matches "com.apify/apify-mcp-server"
const nameParts = entry.fullName.split("/");
if (nameParts.length > 1 && nameParts[1]) {
// Check if it matches the second part (after the slash)
const secondPart = nameParts[1]
.replace("-mcp-server", "")
.replace("-mcp", "");
if (secondPart === serverNameLower) {
return true;
}
}
// Check if serverName matches the displayName ignoring case
return entry.displayName === serverNameLower;
});
const serverLabel = registryEntry
? `[${serverName}](${`https://github.com/mcp/${registryEntry.name}`})`
: serverName;
return `${serverLabel}
${installBadgeUrls}`;
})
.join("
");
}
/**
* Generate the agents section with a table of all agents
* @param {string} agentsDir - Directory path
* @param {{ name: string, displayName: string }[]} registryNames - Pre-loaded MCP registry names
*/
function generateAgentsSection(agentsDir, registryNames = []) {
return generateUnifiedModeSection({
dir: agentsDir,
extension: ".agent.md",
linkPrefix: "agents",
badgeType: "agent",
includeMcpServers: true,
sectionTemplate: TEMPLATES.agentsSection,
usageTemplate: TEMPLATES.agentsUsage,
registryNames,
});
}
/**
* Generate the skills section with a table of all skills
*/
function generateSkillsSection(skillsDir) {
if (!fs.existsSync(skillsDir)) {
console.log(`Skills directory does not exist: ${skillsDir}`);
return "";
}
// Get all skill folders (directories)
const skillFolders = fs.readdirSync(skillsDir).filter((file) => {
const filePath = path.join(skillsDir, file);
return fs.statSync(filePath).isDirectory();
});
// Parse each skill folder
const skillEntries = skillFolders
.map((folder) => {
const skillPath = path.join(skillsDir, folder);
const metadata = parseSkillMetadata(skillPath);
if (!metadata) return null;
return {
folder,
name: metadata.name,
description: metadata.description,
assets: metadata.assets,
};
})
.filter((entry) => entry !== null)
.sort((a, b) => a.name.localeCompare(b.name));
console.log(`Found ${skillEntries.length} skill(s)`);
if (skillEntries.length === 0) {
return "";
}
// Create table header
let content =
"| Name | Description | Bundled Assets |\n| ---- | ----------- | -------------- |\n";
// Generate table rows for each skill
for (const skill of skillEntries) {
const link = `../skills/${skill.folder}/SKILL.md`;
const assetsList =
skill.assets.length > 0
? skill.assets.map((a) => `\`${a}\``).join("
")
: "None";
content += `| [${skill.name}](${link}) | ${formatTableCell(
skill.description
)} | ${assetsList} |\n`;
}
return `${TEMPLATES.skillsSection}\n${TEMPLATES.skillsUsage}\n\n${content}`;
}
/**
* Unified generator for agents (future consolidation)
* @param {Object} cfg
* @param {string} cfg.dir - Directory path
* @param {string} cfg.extension - File extension to match (e.g. .agent.md, .agent.md)
* @param {string} cfg.linkPrefix - Link prefix folder name
* @param {string} cfg.badgeType - Badge key (mode, agent)
* @param {boolean} cfg.includeMcpServers - Whether to include MCP server column
* @param {string} cfg.sectionTemplate - Section heading template
* @param {string} cfg.usageTemplate - Usage subheading template
* @param {{ name: string, displayName: string }[]} cfg.registryNames - Pre-loaded MCP registry names
*/
function generateUnifiedModeSection(cfg) {
const {
dir,
extension,
linkPrefix,
badgeType,
includeMcpServers,
sectionTemplate,
usageTemplate,
registryNames = [],
} = cfg;
if (!fs.existsSync(dir)) {
console.log(`Directory missing for unified mode section: ${dir}`);
return "";
}
const files = fs.readdirSync(dir).filter((f) => f.endsWith(extension));
const entries = files.map((file) => {
const filePath = path.join(dir, file);
return { file, filePath, title: extractTitle(filePath) };
});
entries.sort((a, b) => a.title.localeCompare(b.title));
console.log(
`Unified mode generator: ${entries.length} files for extension ${extension}`
);
if (entries.length === 0) return "";
let header = "| Title | Description |";
if (includeMcpServers) header += " MCP Servers |";
let separator = "| ----- | ----------- |";
if (includeMcpServers) separator += " ----------- |";
let content = `${header}\n${separator}\n`;
for (const { file, filePath, title } of entries) {
const link = encodeURI(`${linkPrefix}/${file}`);
const description = extractDescription(filePath);
const badges = makeBadges(link, badgeType);
let mcpServerCell = "";
if (includeMcpServers) {
const servers = extractMcpServerConfigs(filePath);
mcpServerCell = generateMcpServerLinks(servers, registryNames);
}
const descCell =
description && description !== "null" ? formatTableCell(description) : "";
if (includeMcpServers) {
content += `| [${title}](../${link})
${badges} | ${descCell} | ${mcpServerCell} |\n`;
} else {
content += `| [${title}](../${link})
${badges} | ${descCell} |\n`;
}
}
return `${sectionTemplate}\n${usageTemplate}\n\n${content}`;
}
/**
* Generate the collections section with a table of all collections
*/
function generateCollectionsSection(collectionsDir) {
// Check if collections directory exists, create it if it doesn't
if (!fs.existsSync(collectionsDir)) {
console.log("Collections directory does not exist, creating it...");
fs.mkdirSync(collectionsDir, { recursive: true });
}
// Get all collection files
const collectionFiles = fs
.readdirSync(collectionsDir)
.filter((file) => file.endsWith(".collection.yml"));
// Map collection files to objects with name for sorting
const collectionEntries = collectionFiles
.map((file) => {
const filePath = path.join(collectionsDir, file);
const collection = parseCollectionYaml(filePath);
if (!collection) {
console.warn(`Failed to parse collection: ${file}`);
return null;
}
const collectionId =
collection.id || path.basename(file, ".collection.yml");
const name = collection.name || collectionId;
const isFeatured = collection.display?.featured === true;
return { file, filePath, collection, collectionId, name, isFeatured };
})
.filter((entry) => entry !== null); // Remove failed parses
// Separate featured and regular collections
const featuredCollections = collectionEntries.filter(
(entry) => entry.isFeatured
);
const regularCollections = collectionEntries.filter(
(entry) => !entry.isFeatured
);
// Sort each group alphabetically by name
featuredCollections.sort((a, b) => a.name.localeCompare(b.name));
regularCollections.sort((a, b) => a.name.localeCompare(b.name));
// Combine: featured first, then regular
const sortedEntries = [...featuredCollections, ...regularCollections];
console.log(
`Found ${collectionEntries.length} collection files (${featuredCollections.length} featured)`
);
// If no collections, return empty string
if (sortedEntries.length === 0) {
return "";
}
// Create table header
let collectionsContent =
"| Name | Description | Items | Tags |\n| ---- | ----------- | ----- | ---- |\n";
// Generate table rows for each collection file
for (const entry of sortedEntries) {
const { collection, collectionId, name, isFeatured } = entry;
const description = formatTableCell(
collection.description || "No description"
);
const itemCount = collection.items ? collection.items.length : 0;
const tags = collection.tags ? collection.tags.join(", ") : "";
const link = `../collections/${collectionId}.md`;
const displayName = isFeatured ? `⭐ ${name}` : name;
collectionsContent += `| [${displayName}](${link}) | ${description} | ${itemCount} items | ${tags} |\n`;
}
return `${TEMPLATES.collectionsSection}\n${TEMPLATES.collectionsUsage}\n\n${collectionsContent}`;
}
/**
* Generate the featured collections section for the main README
*/
function generateFeaturedCollectionsSection(collectionsDir) {
// Check if collections directory exists
if (!fs.existsSync(collectionsDir)) {
return "";
}
// Get all collection files
const collectionFiles = fs
.readdirSync(collectionsDir)
.filter((file) => file.endsWith(".collection.yml"));
// Map collection files to objects with name for sorting, filter for featured
const featuredCollections = collectionFiles
.map((file) => {
const filePath = path.join(collectionsDir, file);
return safeFileOperation(
() => {
const collection = parseCollectionYaml(filePath);
if (!collection) return null;
// Only include collections with featured: true
if (!collection.display?.featured) return null;
const collectionId =
collection.id || path.basename(file, ".collection.yml");
const name = collection.name || collectionId;
const description = formatTableCell(
collection.description || "No description"
);
const tags = collection.tags ? collection.tags.join(", ") : "";
const itemCount = collection.items ? collection.items.length : 0;
return {
file,
collection,
collectionId,
name,
description,
tags,
itemCount,
};
},
filePath,
null
);
})
.filter((entry) => entry !== null); // Remove non-featured and failed parses
// Sort by name alphabetically
featuredCollections.sort((a, b) => a.name.localeCompare(b.name));
console.log(`Found ${featuredCollections.length} featured collection(s)`);
// If no featured collections, return empty string
if (featuredCollections.length === 0) {
return "";
}
// Create table header
let featuredContent =
"| Name | Description | Items | Tags |\n| ---- | ----------- | ----- | ---- |\n";
// Generate table rows for each featured collection
for (const entry of featuredCollections) {
const { collectionId, name, description, tags, itemCount } = entry;
const readmeLink = `collections/${collectionId}.md`;
featuredContent += `| [${name}](${readmeLink}) | ${description} | ${itemCount} items | ${tags} |\n`;
}
return `${TEMPLATES.featuredCollectionsSection}\n\n${featuredContent}`;
}
/**
* Generate individual collection README file
* @param {Object} collection - Collection object
* @param {string} collectionId - Collection ID
* @param {{ name: string, displayName: string }[]} registryNames - Pre-loaded MCP registry names
*/
function generateCollectionReadme(
collection,
collectionId,
registryNames = []
) {
if (!collection || !collection.items) {
return `# ${collectionId}\n\nCollection not found or invalid.`;
}
const name = collection.name || collectionId;
const description = collection.description || "No description provided.";
const tags = collection.tags ? collection.tags.join(", ") : "None";
let content = `# ${name}\n\n${description}\n\n`;
if (collection.tags && collection.tags.length > 0) {
content += `**Tags:** ${tags}\n\n`;
}
content += `## Items in this Collection\n\n`;
// Check if collection has any agents to determine table structure (future: chatmodes may migrate)
const hasAgents = collection.items.some((item) => item.kind === "agent");
// Generate appropriate table header
if (hasAgents) {
content += `| Title | Type | Description | MCP Servers |\n| ----- | ---- | ----------- | ----------- |\n`;
} else {
content += `| Title | Type | Description |\n| ----- | ---- | ----------- |\n`;
}
let collectionUsageHeader = "## Collection Usage\n\n";
let collectionUsageContent = [];
// Sort items based on display.ordering setting
const items = [...collection.items];
if (collection.display?.ordering === "alpha") {
items.sort((a, b) => {
const titleA = extractTitle(path.join(ROOT_FOLDER, a.path));
const titleB = extractTitle(path.join(ROOT_FOLDER, b.path));
return titleA.localeCompare(titleB);
});
}
for (const item of items) {
const filePath = path.join(ROOT_FOLDER, item.path);
const title = extractTitle(filePath);
const description = extractDescription(filePath) || "No description";
const typeDisplay =
item.kind === "instruction"
? "Instruction"
: item.kind === "agent"
? "Agent"
: item.kind === "skill"
? "Skill"
: "Prompt";
const link = `../${item.path}`;
// Create install badges for each item (skills don't use chat install badges)
const badgeType =
item.kind === "instruction"
? "instructions"
: item.kind === "agent"
? "agent"
: item.kind === "skill"
? null
: "prompt";
const badges = badgeType ? makeBadges(item.path, badgeType) : "";
const usageDescription = item.usage
? `${description} [see usage](#${title
.replace(/\s+/g, "-")
.toLowerCase()})`
: description;
// Generate MCP server column if collection has agents
content += buildCollectionRow({
hasAgents,
title,
link,
badges,
typeDisplay,
usageDescription,
filePath,
kind: item.kind,
registryNames,
});
// Generate Usage section for each collection
if (item.usage && item.usage.trim()) {
collectionUsageContent.push(
`### ${title}\n\n${item.usage.trim()}\n\n---\n\n`
);
}
}
// Append the usage section if any items had usage defined
if (collectionUsageContent.length > 0) {
content += `\n${collectionUsageHeader}${collectionUsageContent.join("")}`;
} else if (collection.display?.show_badge) {
content += "\n---\n";
}
// Optional badge note at the end if show_badge is true
if (collection.display?.show_badge) {
content += `*This collection includes ${items.length} curated items for **${name}**.*`;
}
return content;
}
/**
* Build a single markdown table row for a collection item.
* Handles optional MCP server column when agents are present.
*/
function buildCollectionRow({
hasAgents,
title,
link,
badges,
typeDisplay,
usageDescription,
filePath,
kind,
registryNames = [],
}) {
const titleCell = badges
? `[${title}](${link})
${badges}`
: `[${title}](${link})`;
// Ensure description is table-safe
const safeUsage = formatTableCell(usageDescription);
if (hasAgents) {
// Only agents currently have MCP servers;
const mcpServers =
kind === "agent" ? extractMcpServerConfigs(filePath) : [];
const mcpServerCell =
mcpServers.length > 0
? generateMcpServerLinks(mcpServers, registryNames)
: "";
return `| ${titleCell} | ${typeDisplay} | ${safeUsage} | ${mcpServerCell} |\n`;
}
return `| ${titleCell} | ${typeDisplay} | ${safeUsage} |\n`;
}
// Utility: write file only if content changed
function writeFileIfChanged(filePath, content) {
const exists = fs.existsSync(filePath);
if (exists) {
const original = fs.readFileSync(filePath, "utf8");
if (original === content) {
console.log(
`${path.basename(filePath)} is already up to date. No changes needed.`
);
return;
}
}
fs.writeFileSync(filePath, content);
console.log(
`${path.basename(filePath)} ${exists ? "updated" : "created"} successfully!`
);
}
// Build per-category README content using existing generators, upgrading headings to H1
function buildCategoryReadme(
sectionBuilder,
dirPath,
headerLine,
usageLine,
registryNames = []
) {
const section = sectionBuilder(dirPath, registryNames);
if (section && section.trim()) {
// Upgrade the first markdown heading level from ## to # for standalone README files
return section.replace(/^##\s/m, "# ");
}
// Fallback content when no entries are found
return `${headerLine}\n\n${usageLine}\n\n_No entries found yet._`;
}
// Main execution wrapped in async function
async function main() {
try {
console.log("Generating category README files...");
// Load MCP registry names once at the beginning
const registryNames = await loadMcpRegistryNames();
// Compose headers for standalone files by converting section headers to H1
const instructionsHeader = TEMPLATES.instructionsSection.replace(
/^##\s/m,
"# "
);
const promptsHeader = TEMPLATES.promptsSection.replace(/^##\s/m, "# ");
const agentsHeader = TEMPLATES.agentsSection.replace(/^##\s/m, "# ");
const skillsHeader = TEMPLATES.skillsSection.replace(/^##\s/m, "# ");
const collectionsHeader = TEMPLATES.collectionsSection.replace(
/^##\s/m,
"# "
);
const instructionsReadme = buildCategoryReadme(
generateInstructionsSection,
INSTRUCTIONS_DIR,
instructionsHeader,
TEMPLATES.instructionsUsage,
registryNames
);
const promptsReadme = buildCategoryReadme(
generatePromptsSection,
PROMPTS_DIR,
promptsHeader,
TEMPLATES.promptsUsage,
registryNames
);
// Generate agents README
const agentsReadme = buildCategoryReadme(
generateAgentsSection,
AGENTS_DIR,
agentsHeader,
TEMPLATES.agentsUsage,
registryNames
);
// Generate skills README
const skillsReadme = buildCategoryReadme(
generateSkillsSection,
SKILLS_DIR,
skillsHeader,
TEMPLATES.skillsUsage,
registryNames
);
// Generate collections README
const collectionsReadme = buildCategoryReadme(
generateCollectionsSection,
COLLECTIONS_DIR,
collectionsHeader,
TEMPLATES.collectionsUsage,
registryNames
);
// Ensure docs directory exists for category outputs
if (!fs.existsSync(DOCS_DIR)) {
fs.mkdirSync(DOCS_DIR, { recursive: true });
}
// Write category outputs into docs folder
writeFileIfChanged(
path.join(DOCS_DIR, "README.instructions.md"),
instructionsReadme
);
writeFileIfChanged(path.join(DOCS_DIR, "README.prompts.md"), promptsReadme);
writeFileIfChanged(path.join(DOCS_DIR, "README.agents.md"), agentsReadme);
writeFileIfChanged(path.join(DOCS_DIR, "README.skills.md"), skillsReadme);
writeFileIfChanged(
path.join(DOCS_DIR, "README.collections.md"),
collectionsReadme
);
// Generate individual collection README files
if (fs.existsSync(COLLECTIONS_DIR)) {
console.log("Generating individual collection README files...");
const collectionFiles = fs
.readdirSync(COLLECTIONS_DIR)
.filter((file) => file.endsWith(".collection.yml"));
for (const file of collectionFiles) {
const filePath = path.join(COLLECTIONS_DIR, file);
const collection = parseCollectionYaml(filePath);
if (collection) {
const collectionId =
collection.id || path.basename(file, ".collection.yml");
const readmeContent = generateCollectionReadme(
collection,
collectionId,
registryNames
);
const readmeFile = path.join(COLLECTIONS_DIR, `${collectionId}.md`);
writeFileIfChanged(readmeFile, readmeContent);
}
}
}
// Generate featured collections section and update main README.md
console.log("Updating main README.md with featured collections...");
const featuredSection = generateFeaturedCollectionsSection(COLLECTIONS_DIR);
if (featuredSection) {
const mainReadmePath = path.join(ROOT_FOLDER, "README.md");
if (fs.existsSync(mainReadmePath)) {
let readmeContent = fs.readFileSync(mainReadmePath, "utf8");
// Define markers to identify where to insert the featured collections
const startMarker = "## 🌟 Featured Collections";
const endMarker = "## MCP Server";
// Check if the section already exists
const startIndex = readmeContent.indexOf(startMarker);
if (startIndex !== -1) {
// Section exists, replace it
const endIndex = readmeContent.indexOf(endMarker, startIndex);
if (endIndex !== -1) {
// Replace the existing section
const beforeSection = readmeContent.substring(0, startIndex);
const afterSection = readmeContent.substring(endIndex);
readmeContent =
beforeSection + featuredSection + "\n\n" + afterSection;
}
} else {
// Section doesn't exist, insert it before "## MCP Server"
const mcpIndex = readmeContent.indexOf(endMarker);
if (mcpIndex !== -1) {
const beforeMcp = readmeContent.substring(0, mcpIndex);
const afterMcp = readmeContent.substring(mcpIndex);
readmeContent = beforeMcp + featuredSection + "\n\n" + afterMcp;
}
}
writeFileIfChanged(mainReadmePath, readmeContent);
console.log("Main README.md updated with featured collections");
} else {
console.warn(
"README.md not found, skipping featured collections update"
);
}
} else {
console.log("No featured collections found to add to README.md");
}
} catch (error) {
console.error(`Error generating category README files: ${error.message}`);
console.error(error.stack);
process.exit(1);
}
}
// Run the main function
main().catch((error) => {
console.error(`Fatal error: ${error.message}`);
console.error(error.stack);
process.exit(1);
});