* initial prototype of partners collection with featured collection support

* Starting to add the partners

* Preparing the repo for how the custom agents will work

* moving some files around

* Moving a bunch of stuff around to make the file easier to read

* improving the front matter parsing by using a real library

* Some verbage updates

* some more verbage

* Fixing spelling mistake

* tweaking badges

* Updating contributing guide to be correct

* updating casing to match product

* More agents

* Better handling link to mcp registry

* links to install mcp servers fixed up

* Updating collection tags

* writing the mcp registry url out properly

* Adding custom agents for C# and WinForms

Expert custom agents to improve your experience when working with C# and WinForms in Copilot

* Adding to agents readme

* Adding PagerDuty agent

* Fixing description for terraform agent

* Adding custom agents to the README usage

* Removing the button to make the links more obvious

* docs: relocate category READMEs to /docs and update generation + internal links

* Updating prompts for new path

* formatting

---------

Co-authored-by: Chris Patterson <chrispat@github.com>
This commit is contained in:
Aaron Powell
2025-10-29 06:07:13 +11:00
committed by GitHub
parent f4533e683c
commit 56d7ce73a0
55 changed files with 5683 additions and 1110 deletions

132
eng/constants.js Normal file
View File

@@ -0,0 +1,132 @@
const path = require("path");
// Template sections for the README
const TEMPLATES = {
instructionsSection: `## 📋 Custom Instructions
Team and project-specific instructions to enhance GitHub Copilot's behavior for specific technologies and coding practices.`,
instructionsUsage: `### How to Use Custom Instructions
**To Install:**
- Click the **VS Code** or **VS Code Insiders** install button for the instruction you want to use
- Download the \`*.instructions.md\` file and manually add it to your project's instruction collection
**To Use/Apply:**
- Copy these instructions to your \`.github/copilot-instructions.md\` file in your workspace
- Create task-specific \`.github/.instructions.md\` files in your workspace's \`.github/instructions\` folder
- Instructions automatically apply to Copilot behavior once installed in your workspace`,
promptsSection: `## 🎯 Reusable Prompts
Ready-to-use prompt templates for specific development scenarios and tasks, defining prompt text with a specific mode, model, and available set of tools.`,
promptsUsage: `### How to Use Reusable Prompts
**To Install:**
- Click the **VS Code** or **VS Code Insiders** install button for the prompt you want to use
- Download the \`*.prompt.md\` file and manually add it to your prompt collection
**To Run/Execute:**
- Use \`/prompt-name\` in VS Code chat after installation
- Run the \`Chat: Run Prompt\` command from the Command Palette
- Hit the run button while you have a prompt file open in VS Code`,
chatmodesSection: `## 💭 Custom Chat Modes
Custom chat modes define specific behaviors and tools for GitHub Copilot Chat, enabling enhanced context-aware assistance for particular tasks or workflows.`,
chatmodesUsage: `### How to Use Custom Chat Modes
**To Install:**
- Click the **VS Code** or **VS Code Insiders** install button for the chat mode you want to use
- Download the \`*.chatmode.md\` file and manually install it in VS Code using the Command Palette
**To Activate/Use:**
- Import the chat mode configuration into your VS Code settings
- Access the installed chat modes through the VS Code Chat interface
- Select the desired chat mode from the available options in VS Code Chat`,
collectionsSection: `## 📦 Collections
Curated collections of related prompts, instructions, and chat modes organized around specific themes, workflows, or use cases.`,
collectionsUsage: `### How to Use Collections
**Browse Collections:**
- ⭐ Featured collections are highlighted and appear at the top of the list
- Explore themed collections that group related customizations
- Each collection includes prompts, instructions, and chat modes for specific workflows
- Collections make it easy to adopt comprehensive toolkits for particular scenarios
**Install Items:**
- Click install buttons for individual items within collections
- Or browse to the individual files to copy content manually
- Collections help you discover related customizations you might have missed`,
featuredCollectionsSection: `## 🌟 Featured Collections
Discover our curated collections of prompts, instructions, and chat modes organized around specific themes and workflows.`,
agentsSection: `## 🤖 Custom Agents
Custom agents for GitHub Copilot, making it easy for users and organizations to "specialize" their Copilot coding agent (CCA) through simple file-based configuration.`,
agentsUsage: `### How to Use Custom Agents
**To Install:**
- Click the **VS Code** or **VS Code Insiders** install button for the agent you want to use
- Download the \`*.agent.md\` file and add it to your repository
**MCP Server Setup:**
- Each agent may require one or more MCP servers to function
- Click the MCP server to view it on the GitHub MCP registry
- Follow the guide on how to add the MCP server to your repository
**To Activate/Use:**
- Access installed agents through the VS Code Chat interface, assign them in CCA, or through Copilot CLI (coming soon)
- Agents will have access to tools from configured MCP servers
- Follow agent-specific instructions for optimal usage`,
};
const vscodeInstallImage =
"https://img.shields.io/badge/VS_Code-Install-0098FF?style=flat-square&logo=visualstudiocode&logoColor=white";
const vscodeInsidersInstallImage =
"https://img.shields.io/badge/VS_Code_Insiders-Install-24bfa5?style=flat-square&logo=visualstudiocode&logoColor=white";
const repoBaseUrl =
"https://raw.githubusercontent.com/github/awesome-copilot/main";
const AKA_INSTALL_URLS = {
instructions: "https://aka.ms/awesome-copilot/install/instructions",
prompt: "https://aka.ms/awesome-copilot/install/prompt",
mode: "https://aka.ms/awesome-copilot/install/chatmode",
agent: "https://aka.ms/awesome-copilot/install/agent",
};
const ROOT_FOLDER = path.join(__dirname, "..");
const INSTRUCTIONS_DIR = path.join(ROOT_FOLDER, "instructions");
const PROMPTS_DIR = path.join(ROOT_FOLDER, "prompts");
const CHATMODES_DIR = path.join(ROOT_FOLDER, "chatmodes");
const AGENTS_DIR = path.join(ROOT_FOLDER, "agents");
const COLLECTIONS_DIR = path.join(ROOT_FOLDER, "collections");
const MAX_COLLECTION_ITEMS = 50;
const DOCS_DIR = path.join(ROOT_FOLDER, "docs");
module.exports = {
TEMPLATES,
vscodeInstallImage,
vscodeInsidersInstallImage,
repoBaseUrl,
AKA_INSTALL_URLS,
ROOT_FOLDER,
INSTRUCTIONS_DIR,
PROMPTS_DIR,
CHATMODES_DIR,
AGENTS_DIR,
COLLECTIONS_DIR,
MAX_COLLECTION_ITEMS,
DOCS_DIR,
};

182
eng/create-collection.js Executable file
View File

@@ -0,0 +1,182 @@
#!/usr/bin/env node
const fs = require("fs");
const path = require("path");
const readline = require("readline");
const rl = readline.createInterface({
input: process.stdin,
output: process.stdout,
});
function prompt(question) {
return new Promise((resolve) => {
rl.question(question, resolve);
});
}
function parseArgs() {
const args = process.argv.slice(2);
const out = { id: undefined, tags: undefined };
// simple long/short option parsing
for (let i = 0; i < args.length; i++) {
const a = args[i];
if (a === "--id" || a === "-i") {
out.id = args[i + 1];
i++;
} else if (a.startsWith("--id=")) {
out.id = a.split("=")[1];
} else if (a === "--tags" || a === "-t") {
out.tags = args[i + 1];
i++;
} else if (a.startsWith("--tags=")) {
out.tags = a.split("=")[1];
} else if (!a.startsWith("-") && !out.id) {
// first positional -> id
out.id = a;
} else if (!a.startsWith("-") && out.id && !out.tags) {
// second positional -> tags
out.tags = a;
}
}
// normalize tags to string (comma separated) or undefined
if (Array.isArray(out.tags)) {
out.tags = out.tags.join(",");
}
return out;
}
async function createCollectionTemplate() {
try {
console.log("🎯 Collection Creator");
console.log("This tool will help you create a new collection manifest.\n");
// Parse CLI args and fall back to interactive prompts when missing
const parsed = parseArgs();
// Get collection ID
let collectionId = parsed.id;
if (!collectionId) {
collectionId = await prompt("Collection ID (lowercase, hyphens only): ");
}
// Validate collection ID format
if (!collectionId) {
console.error("❌ Collection ID is required");
process.exit(1);
}
if (!/^[a-z0-9-]+$/.test(collectionId)) {
console.error(
"❌ Collection ID must contain only lowercase letters, numbers, and hyphens"
);
process.exit(1);
}
const collectionsDir = path.join(__dirname, "collections");
const filePath = path.join(
collectionsDir,
`${collectionId}.collection.yml`
);
// Check if file already exists
if (fs.existsSync(filePath)) {
console.log(
`⚠️ Collection ${collectionId} already exists at ${filePath}`
);
console.log("💡 Please edit that file instead or choose a different ID.");
process.exit(1);
}
// Ensure collections directory exists
if (!fs.existsSync(collectionsDir)) {
fs.mkdirSync(collectionsDir, { recursive: true });
}
// Get collection name
const defaultName = collectionId
.split("-")
.map((word) => word.charAt(0).toUpperCase() + word.slice(1))
.join(" ");
let collectionName = await prompt(
`Collection name (default: ${defaultName}): `
);
if (!collectionName.trim()) {
collectionName = defaultName;
}
// Get description
const defaultDescription = `A collection of related prompts, instructions, and chat modes for ${collectionName.toLowerCase()}.`;
let description = await prompt(
`Description (default: ${defaultDescription}): `
);
if (!description.trim()) {
description = defaultDescription;
}
// Get tags (from CLI or prompt)
let tags = [];
let tagInput = parsed.tags;
if (!tagInput) {
tagInput = await prompt(
"Tags (comma-separated, or press Enter for defaults): "
);
}
if (tagInput && tagInput.toString().trim()) {
tags = tagInput
.toString()
.split(",")
.map((tag) => tag.trim())
.filter((tag) => tag);
} else {
// Generate some default tags from the collection ID
tags = collectionId.split("-").slice(0, 3);
}
// Template content
const template = `id: ${collectionId}
name: ${collectionName}
description: ${description}
tags: [${tags.join(", ")}]
items:
# Add your collection items here
# Example:
# - path: prompts/example.prompt.md
# kind: prompt
# - path: instructions/example.instructions.md
# kind: instruction
# - path: chatmodes/example.chatmode.md
# kind: chat-mode
# - path: agents/example.agent.md
# kind: agent
# usage: |
# This agent requires the example MCP server to be installed.
# Configure any required environment variables (e.g., EXAMPLE_API_KEY).
display:
ordering: alpha # or "manual" to preserve the order above
show_badge: false # set to true to show collection badge on items
`;
fs.writeFileSync(filePath, template);
console.log(`✅ Created collection template: ${filePath}`);
console.log("\n📝 Next steps:");
console.log("1. Edit the collection manifest to add your items");
console.log("2. Update the name, description, and tags as needed");
console.log("3. Run 'npm run validate:collections' to validate");
console.log("4. Run 'npm start' to generate documentation");
console.log("\n📄 Collection template contents:");
console.log(template);
} catch (error) {
console.error(`❌ Error creating collection template: ${error.message}`);
process.exit(1);
} finally {
rl.close();
}
}
// Run the interactive creation process
createCollectionTemplate();

1137
eng/github-mcp-registry.json Normal file

File diff suppressed because it is too large Load Diff

970
eng/update-readme.js Executable file
View File

@@ -0,0 +1,970 @@
#!/usr/bin/env node
const fs = require("fs");
const path = require("path");
const {
parseCollectionYaml,
extractMcpServers,
extractMcpServerConfigs,
parseFrontmatter,
} = require("./yaml-parser");
const {
TEMPLATES,
AKA_INSTALL_URLS,
repoBaseUrl,
vscodeInstallImage,
vscodeInsidersInstallImage,
ROOT_FOLDER,
PROMPTS_DIR,
CHATMODES_DIR,
AGENTS_DIR,
COLLECTIONS_DIR,
INSTRUCTIONS_DIR,
DOCS_DIR,
} = require("./constants");
// Cache of MCP registry server names (lower-cased) loaded from github-mcp-registry.json
let MCP_REGISTRY_SET = null;
/**
* Loads and caches the set of MCP registry server display names (lowercased).
*
* Behavior:
* - If a cached set already exists (MCP_REGISTRY_SET), it is returned immediately.
* - Attempts to read a JSON registry file named "github-mcp-registry.json" from the
* same directory as this script.
* - Safely handles missing file or malformed JSON by returning an empty Set.
* - Extracts server display names from: json.payload.mcpRegistryRoute.serversData.servers
* - Normalizes names to lowercase and stores them in a Set for O(1) membership checks.
*
* Side Effects:
* - Mutates the module-scoped variable MCP_REGISTRY_SET.
* - Logs a warning to console if reading or parsing the registry fails.
*
* @returns {{ name: string, displayName: string }[]} A Set of lowercased server display names. May be empty if
* the registry file is absent, unreadable, or malformed.
*
* @throws {none} All errors are caught internally; failures result in an empty Set.
*/
function loadMcpRegistryNames() {
if (MCP_REGISTRY_SET) return MCP_REGISTRY_SET;
try {
const registryPath = path.join(__dirname, "github-mcp-registry.json");
if (!fs.existsSync(registryPath)) {
MCP_REGISTRY_SET = [];
return MCP_REGISTRY_SET;
}
const raw = fs.readFileSync(registryPath, "utf8");
const json = JSON.parse(raw);
const servers = json?.payload?.mcpRegistryRoute?.serversData?.servers || [];
MCP_REGISTRY_SET = servers.map((s) => ({
name: s.name,
displayName: s.display_name.toLowerCase(),
}));
} catch (e) {
console.warn(`Failed to load MCP registry: ${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/chatmode/instructions files, look for heading after frontmatter
if (
filePath.includes(".prompt.md") ||
filePath.includes(".chatmode.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(".chatmode.md")
? ".chatmode.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
);
}
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}/${link}`
)}`;
return `[![Install in VS Code](${vscodeInstallImage})](${vscodeUrl})<br />[![Install in VS Code Insiders](${vscodeInsidersInstallImage})](${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
instructionsContent += `| [${title}](${link})<br />${badges} | ${customDescription} |\n`;
} else {
// Fallback to the default approach - use last word of title for description, removing trailing 's' if present
const topic = title.split(" ").pop().replace(/s$/, "");
instructionsContent += `| [${title}](${link})<br />${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})<br />${badges} | ${customDescription} |\n`;
} else {
promptsContent += `| [${title}](${link})<br />${badges} | | |\n`;
}
}
return `${TEMPLATES.promptsSection}\n${TEMPLATES.promptsUsage}\n\n${promptsContent}`;
}
/**
* Generate the chat modes section with a table of all chat modes
*/
function generateChatModesSection(chatmodesDir) {
return generateUnifiedModeSection({
dir: chatmodesDir,
extension: ".chatmode.md",
linkPrefix: "chatmodes",
badgeType: "mode",
includeMcpServers: false,
sectionTemplate: TEMPLATES.chatmodesSection,
usageTemplate: TEMPLATES.chatmodesUsage,
});
}
/**
* Generate MCP server links for an agent
* @param {string[]} servers - Array of MCP server names
* @returns {string} - Formatted MCP server links with badges
*/
function generateMcpServerLinks(servers) {
if (!servers || servers.length === 0) {
return "";
}
const badges = [
{
type: "vscode",
url: "https://img.shields.io/badge/Install-VS_Code-0098FF?style=flat-square",
badgeUrl: (serverName) =>
`https://aka.ms/awesome-copilot/install/mcp-vscode?vscode:mcp/by-name/${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`,
},
];
const registryNames = loadMcpRegistryNames();
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 = [
`[![Install MCP](${badges[0].url})](https://aka.ms/awesome-copilot/install/mcp-vscode?name=${serverName}&config=${encodedConfig})`,
`[![Install MCP](${badges[1].url})](https://aka.ms/awesome-copilot/install/mcp-vscodeinsiders?name=${serverName}&config=${encodedConfig})`,
`[![Install MCP](${badges[2].url})](https://aka.ms/awesome-copilot/install/mcp-visualstudio/mcp-install?${encodedConfig})`,
].join("<br />");
const registryEntry = registryNames.find(
(entry) => entry.displayName === serverName.toLowerCase()
);
const serverLabel = registryEntry
? `[${serverName}](${`https://github.com/mcp/${registryEntry.name}`})`
: serverName;
return `${serverLabel}<br />${installBadgeUrls}`;
})
.join("<br />");
}
/**
* Generate the agents section with a table of all agents
*/
function generateAgentsSection(agentsDir) {
return generateUnifiedModeSection({
dir: agentsDir,
extension: ".agent.md",
linkPrefix: "agents",
badgeType: "agent",
includeMcpServers: true,
sectionTemplate: TEMPLATES.agentsSection,
usageTemplate: TEMPLATES.agentsUsage,
});
}
/**
* Unified generator for chat modes & agents (future consolidation)
* @param {Object} cfg
* @param {string} cfg.dir - Directory path
* @param {string} cfg.extension - File extension to match (e.g. .chatmode.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
*/
function generateUnifiedModeSection(cfg) {
const {
dir,
extension,
linkPrefix,
badgeType,
includeMcpServers,
sectionTemplate,
usageTemplate,
} = 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}/${file}`);
const description = extractDescription(filePath);
const badges = makeBadges(link, badgeType);
let mcpServerCell = "";
if (includeMcpServers) {
const servers = extractMcpServerConfigs(filePath);
mcpServerCell = generateMcpServerLinks(servers);
}
if (includeMcpServers) {
content += `| [${title}](${link})<br />${badges} | ${
description && description !== "null" ? description : ""
} | ${mcpServerCell} |\n`;
} else {
content += `| [${title}](${link})<br />${badges} | ${
description && description !== "null" ? description : ""
} |\n`;
}
}
return `${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 = 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 = 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
*/
function generateCollectionReadme(collection, collectionId) {
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:** ${tags}\n\n`;
}
content += `## Items in this Collection\n\n`;
// Check if collection has any agents to determine table structure (future: chatmodes may migrate)
const hasAgents = collection.items.some((item) => item.kind === "agent");
// Generate appropriate table header
if (hasAgents) {
content += `| Title | Type | Description | MCP Servers |\n| ----- | ---- | ----------- | ----------- |\n`;
} else {
content += `| Title | Type | Description |\n| ----- | ---- | ----------- |\n`;
}
let collectionUsageHeader = "## Collection Usage\n\n";
let collectionUsageContent = [];
// Sort items based on display.ordering setting
const items = [...collection.items];
if (collection.display?.ordering === "alpha") {
items.sort((a, b) => {
const titleA = extractTitle(path.join(ROOT_FOLDER, a.path));
const titleB = extractTitle(path.join(ROOT_FOLDER, b.path));
return titleA.localeCompare(titleB);
});
}
for (const item of items) {
const filePath = path.join(ROOT_FOLDER, item.path);
const title = extractTitle(filePath);
const description = extractDescription(filePath) || "No description";
const typeDisplay =
item.kind === "chat-mode"
? "Chat Mode"
: item.kind === "instruction"
? "Instruction"
: item.kind === "agent"
? "Agent"
: "Prompt";
const link = `../${item.path}`;
// Create install badges for each item
const badges = makeBadges(
item.path,
item.kind === "instruction"
? "instructions"
: item.kind === "chat-mode"
? "mode"
: item.kind === "agent"
? "agent"
: "prompt"
);
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,
});
// 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,
}) {
if (hasAgents) {
// Only agents currently have MCP servers; future migration may extend to chat modes.
const mcpServers =
kind === "agent" ? extractMcpServerConfigs(filePath) : [];
const mcpServerCell =
mcpServers.length > 0 ? generateMcpServerLinks(mcpServers) : "";
return `| [${title}](${link})<br />${badges} | ${typeDisplay} | ${usageDescription} | ${mcpServerCell} |\n`;
}
return `| [${title}](${link})<br />${badges} | ${typeDisplay} | ${usageDescription} |\n`;
}
// Utility: write file only if content changed
function writeFileIfChanged(filePath, content) {
const exists = fs.existsSync(filePath);
if (exists) {
const original = fs.readFileSync(filePath, "utf8");
if (original === content) {
console.log(
`${path.basename(filePath)} is already up to date. No changes needed.`
);
return;
}
}
fs.writeFileSync(filePath, content);
console.log(
`${path.basename(filePath)} ${exists ? "updated" : "created"} successfully!`
);
}
// Build per-category README content using existing generators, upgrading headings to H1
function buildCategoryReadme(sectionBuilder, dirPath, headerLine, usageLine) {
const section = sectionBuilder(dirPath);
if (section && section.trim()) {
// Upgrade the first markdown heading level from ## to # for standalone README files
return section.replace(/^##\s/m, "# ");
}
// Fallback content when no entries are found
return `${headerLine}\n\n${usageLine}\n\n_No entries found yet._`;
}
// Main execution
try {
console.log("Generating category README files...");
// 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 chatmodesHeader = TEMPLATES.chatmodesSection.replace(/^##\s/m, "# ");
const agentsHeader = TEMPLATES.agentsSection.replace(/^##\s/m, "# ");
const collectionsHeader = TEMPLATES.collectionsSection.replace(
/^##\s/m,
"# "
);
const instructionsReadme = buildCategoryReadme(
generateInstructionsSection,
INSTRUCTIONS_DIR,
instructionsHeader,
TEMPLATES.instructionsUsage
);
const promptsReadme = buildCategoryReadme(
generatePromptsSection,
PROMPTS_DIR,
promptsHeader,
TEMPLATES.promptsUsage
);
const chatmodesReadme = buildCategoryReadme(
generateChatModesSection,
CHATMODES_DIR,
chatmodesHeader,
TEMPLATES.chatmodesUsage
);
// Generate agents README
const agentsReadme = buildCategoryReadme(
generateAgentsSection,
AGENTS_DIR,
agentsHeader,
TEMPLATES.agentsUsage
);
// Generate collections README
const collectionsReadme = buildCategoryReadme(
generateCollectionsSection,
COLLECTIONS_DIR,
collectionsHeader,
TEMPLATES.collectionsUsage
);
// 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.chatmodes.md"),
chatmodesReadme
);
writeFileIfChanged(path.join(DOCS_DIR, "README.agents.md"), agentsReadme);
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
);
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}`);
process.exit(1);
}

372
eng/validate-collections.js Executable file
View File

@@ -0,0 +1,372 @@
#!/usr/bin/env node
const fs = require("fs");
const path = require("path");
const { parseCollectionYaml, parseFrontmatter } = require("./yaml-parser");
const {
ROOT_FOLDER,
COLLECTIONS_DIR,
MAX_COLLECTION_ITEMS,
} = require("./constants");
// Validation functions
function validateCollectionId(id) {
if (!id || typeof id !== "string") {
return "ID is required and must be a string";
}
if (!/^[a-z0-9-]+$/.test(id)) {
return "ID must contain only lowercase letters, numbers, and hyphens";
}
if (id.length < 1 || id.length > 50) {
return "ID must be between 1 and 50 characters";
}
return null;
}
function validateCollectionName(name) {
if (!name || typeof name !== "string") {
return "Name is required and must be a string";
}
if (name.length < 1 || name.length > 100) {
return "Name must be between 1 and 100 characters";
}
return null;
}
function validateCollectionDescription(description) {
if (!description || typeof description !== "string") {
return "Description is required and must be a string";
}
if (description.length < 1 || description.length > 500) {
return "Description must be between 1 and 500 characters";
}
return null;
}
function validateCollectionTags(tags) {
if (tags && !Array.isArray(tags)) {
return "Tags must be an array";
}
if (tags && tags.length > 10) {
return "Maximum 10 tags allowed";
}
if (tags) {
for (const tag of tags) {
if (typeof tag !== "string") {
return "All tags must be strings";
}
if (!/^[a-z0-9-]+$/.test(tag)) {
return `Tag "${tag}" must contain only lowercase letters, numbers, and hyphens`;
}
if (tag.length < 1 || tag.length > 30) {
return `Tag "${tag}" must be between 1 and 30 characters`;
}
}
}
return null;
}
function validateAgentFile(filePath) {
try {
const agent = parseFrontmatter(filePath);
if (!agent) {
return `Item ${filePath} agent file could not be parsed`;
}
// Validate name field
if (!agent.name || typeof agent.name !== "string") {
return `Item ${filePath} agent must have a 'name' field`;
}
if (agent.name.length < 1 || agent.name.length > 50) {
return `Item ${filePath} agent name must be between 1 and 50 characters`;
}
// Validate description field
if (!agent.description || typeof agent.description !== "string") {
return `Item ${filePath} agent must have a 'description' field`;
}
if (agent.description.length < 1 || agent.description.length > 500) {
return `Item ${filePath} agent description must be between 1 and 500 characters`;
}
// Validate tools field (optional)
if (agent.tools !== undefined && !Array.isArray(agent.tools)) {
return `Item ${filePath} agent 'tools' must be an array`;
}
// Validate mcp-servers field (optional)
if (agent["mcp-servers"]) {
if (
typeof agent["mcp-servers"] !== "object" ||
Array.isArray(agent["mcp-servers"])
) {
return `Item ${filePath} agent 'mcp-servers' must be an object`;
}
// Validate each MCP server configuration
for (const [serverName, serverConfig] of Object.entries(
agent["mcp-servers"]
)) {
if (!serverConfig || typeof serverConfig !== "object") {
return `Item ${filePath} agent MCP server '${serverName}' must be an object`;
}
if (!serverConfig.type || typeof serverConfig.type !== "string") {
return `Item ${filePath} agent MCP server '${serverName}' must have a 'type' field`;
}
// For local type servers, command is required
if (serverConfig.type === "local" && !serverConfig.command) {
return `Item ${filePath} agent MCP server '${serverName}' with type 'local' must have a 'command' field`;
}
// Validate args if present
if (
serverConfig.args !== undefined &&
!Array.isArray(serverConfig.args)
) {
return `Item ${filePath} agent MCP server '${serverName}' 'args' must be an array`;
}
// Validate tools if present
if (
serverConfig.tools !== undefined &&
!Array.isArray(serverConfig.tools)
) {
return `Item ${filePath} agent MCP server '${serverName}' 'tools' must be an array`;
}
// Validate env if present
if (serverConfig.env !== undefined) {
if (
typeof serverConfig.env !== "object" ||
Array.isArray(serverConfig.env)
) {
return `Item ${filePath} agent MCP server '${serverName}' 'env' must be an object`;
}
}
}
}
return null; // All validations passed
} catch (error) {
return `Item ${filePath} agent file validation failed: ${error.message}`;
}
}
function validateCollectionItems(items) {
if (!items || !Array.isArray(items)) {
return "Items is required and must be an array";
}
if (items.length < 1) {
return "At least one item is required";
}
if (items.length > MAX_COLLECTION_ITEMS) {
return `Maximum ${MAX_COLLECTION_ITEMS} items allowed`;
}
for (let i = 0; i < items.length; i++) {
const item = items[i];
if (!item || typeof item !== "object") {
return `Item ${i + 1} must be an object`;
}
if (!item.path || typeof item.path !== "string") {
return `Item ${i + 1} must have a path string`;
}
if (!item.kind || typeof item.kind !== "string") {
return `Item ${i + 1} must have a kind string`;
}
if (!["prompt", "instruction", "chat-mode", "agent"].includes(item.kind)) {
return `Item ${
i + 1
} kind must be one of: prompt, instruction, chat-mode, agent`;
}
// Validate file path exists
const filePath = path.join(ROOT_FOLDER, item.path);
if (!fs.existsSync(filePath)) {
return `Item ${i + 1} file does not exist: ${item.path}`;
}
// Validate path pattern matches kind
if (item.kind === "prompt" && !item.path.endsWith(".prompt.md")) {
return `Item ${
i + 1
} kind is "prompt" but path doesn't end with .prompt.md`;
}
if (
item.kind === "instruction" &&
!item.path.endsWith(".instructions.md")
) {
return `Item ${
i + 1
} kind is "instruction" but path doesn't end with .instructions.md`;
}
if (item.kind === "chat-mode" && !item.path.endsWith(".chatmode.md")) {
return `Item ${
i + 1
} kind is "chat-mode" but path doesn't end with .chatmode.md`;
}
if (item.kind === "agent" && !item.path.endsWith(".agent.md")) {
return `Item ${
i + 1
} kind is "agent" but path doesn't end with .agent.md`;
}
// Validate agent-specific frontmatter
if (item.kind === "agent") {
const agentValidation = validateAgentFile(filePath, i + 1);
if (agentValidation) {
return agentValidation;
}
}
}
return null;
}
function validateCollectionDisplay(display) {
if (display && typeof display !== "object") {
return "Display must be an object";
}
if (display) {
// Normalize ordering and show_badge in case the YAML parser left inline comments
const normalize = (val) => {
if (typeof val !== "string") return val;
// Strip any inline comment starting with '#'
const hashIndex = val.indexOf("#");
if (hashIndex !== -1) {
val = val.substring(0, hashIndex).trim();
}
// Also strip surrounding quotes if present
if (
(val.startsWith('"') && val.endsWith('"')) ||
(val.startsWith("'") && val.endsWith("'"))
) {
val = val.substring(1, val.length - 1);
}
return val.trim();
};
if (display.ordering) {
const normalizedOrdering = normalize(display.ordering);
if (!["manual", "alpha"].includes(normalizedOrdering)) {
return "Display ordering must be 'manual' or 'alpha'";
}
}
if (display.show_badge !== undefined) {
const raw = display.show_badge;
const normalizedBadge = normalize(raw);
// Accept boolean or string boolean values
if (typeof normalizedBadge === "string") {
if (!["true", "false"].includes(normalizedBadge.toLowerCase())) {
return "Display show_badge must be boolean";
}
} else if (typeof normalizedBadge !== "boolean") {
return "Display show_badge must be boolean";
}
}
}
return null;
}
function validateCollectionManifest(collection, filePath) {
const errors = [];
const idError = validateCollectionId(collection.id);
if (idError) errors.push(`ID: ${idError}`);
const nameError = validateCollectionName(collection.name);
if (nameError) errors.push(`Name: ${nameError}`);
const descError = validateCollectionDescription(collection.description);
if (descError) errors.push(`Description: ${descError}`);
const tagsError = validateCollectionTags(collection.tags);
if (tagsError) errors.push(`Tags: ${tagsError}`);
const itemsError = validateCollectionItems(collection.items);
if (itemsError) errors.push(`Items: ${itemsError}`);
const displayError = validateCollectionDisplay(collection.display);
if (displayError) errors.push(`Display: ${displayError}`);
return errors;
}
// Main validation function
function validateCollections() {
if (!fs.existsSync(COLLECTIONS_DIR)) {
console.log("No collections directory found - validation skipped");
return true;
}
const collectionFiles = fs
.readdirSync(COLLECTIONS_DIR)
.filter((file) => file.endsWith(".collection.yml"));
if (collectionFiles.length === 0) {
console.log("No collection files found - validation skipped");
return true;
}
console.log(`Validating ${collectionFiles.length} collection files...`);
let hasErrors = false;
const usedIds = new Set();
for (const file of collectionFiles) {
const filePath = path.join(COLLECTIONS_DIR, file);
console.log(`\nValidating ${file}...`);
const collection = parseCollectionYaml(filePath);
if (!collection) {
console.error(`❌ Failed to parse ${file}`);
hasErrors = true;
continue;
}
// Validate the collection structure
const errors = validateCollectionManifest(collection, filePath);
if (errors.length > 0) {
console.error(`❌ Validation errors in ${file}:`);
errors.forEach((error) => console.error(` - ${error}`));
hasErrors = true;
} else {
console.log(`${file} is valid`);
}
// Check for duplicate IDs
if (collection.id) {
if (usedIds.has(collection.id)) {
console.error(
`❌ Duplicate collection ID "${collection.id}" found in ${file}`
);
hasErrors = true;
} else {
usedIds.add(collection.id);
}
}
}
if (!hasErrors) {
console.log(`\n✅ All ${collectionFiles.length} collections are valid`);
}
return !hasErrors;
}
// Run validation
try {
const isValid = validateCollections();
if (!isValid) {
console.error("\n❌ Collection validation failed");
process.exit(1);
}
console.log("\n🎉 Collection validation passed");
} catch (error) {
console.error(`Error during validation: ${error.message}`);
process.exit(1);
}

147
eng/yaml-parser.js Normal file
View File

@@ -0,0 +1,147 @@
// YAML parser for collection files and frontmatter parsing using vfile-matter
const fs = require("fs");
const yaml = require("js-yaml");
const { VFile } = require("vfile");
const { matter } = require("vfile-matter");
function safeFileOperation(operation, filePath, defaultValue = null) {
try {
return operation();
} catch (error) {
console.error(`Error processing file ${filePath}: ${error.message}`);
return defaultValue;
}
}
/**
* Parse a collection YAML file (.collection.yml)
* Collections are pure YAML files without frontmatter delimiters
* @param {string} filePath - Path to the collection file
* @returns {object|null} Parsed collection object or null on error
*/
function parseCollectionYaml(filePath) {
return safeFileOperation(
() => {
const content = fs.readFileSync(filePath, "utf8");
// Collections are pure YAML files, parse directly with js-yaml
return yaml.load(content, { schema: yaml.JSON_SCHEMA });
},
filePath,
null
);
}
/**
* Parse frontmatter from a markdown file using vfile-matter
* Works with any markdown file that has YAML frontmatter (agents, prompts, chatmodes, instructions)
* @param {string} filePath - Path to the markdown file
* @returns {object|null} Parsed frontmatter object or null on error
*/
function parseFrontmatter(filePath) {
return safeFileOperation(
() => {
const content = fs.readFileSync(filePath, "utf8");
const file = new VFile({ path: filePath, value: content });
// Parse the frontmatter using vfile-matter
matter(file);
// The frontmatter is now available in file.data.matter
const frontmatter = file.data.matter;
// Normalize string fields that can accumulate trailing newlines/spaces
if (frontmatter) {
if (typeof frontmatter.name === "string") {
frontmatter.name = frontmatter.name.replace(/[\r\n]+$/g, "").trim();
}
if (typeof frontmatter.title === "string") {
frontmatter.title = frontmatter.title.replace(/[\r\n]+$/g, "").trim();
}
if (typeof frontmatter.description === "string") {
// Remove only trailing whitespace/newlines; preserve internal formatting
frontmatter.description = frontmatter.description.replace(
/[\s\r\n]+$/g,
""
);
}
}
return frontmatter;
},
filePath,
null
);
}
/**
* Extract agent metadata including MCP server information
* @param {string} filePath - Path to the agent file
* @returns {object|null} Agent metadata object with name, description, tools, and mcp-servers
*/
function extractAgentMetadata(filePath) {
const frontmatter = parseFrontmatter(filePath);
if (!frontmatter) {
return null;
}
return {
name: typeof frontmatter.name === "string" ? frontmatter.name : null,
description:
typeof frontmatter.description === "string"
? frontmatter.description
: null,
tools: frontmatter.tools || [],
mcpServers: frontmatter["mcp-servers"] || {},
};
}
/**
* Extract MCP server names from an agent file
* @param {string} filePath - Path to the agent file
* @returns {string[]} Array of MCP server names
*/
function extractMcpServers(filePath) {
const metadata = extractAgentMetadata(filePath);
if (!metadata || !metadata.mcpServers) {
return [];
}
return Object.keys(metadata.mcpServers);
}
/**
* Extract full MCP server configs from an agent file
* @param {string} filePath - Path to the agent file
* @returns {Array<{name:string,type?:string,command?:string,args?:string[],url?:string,headers?:object}>}
*/
function extractMcpServerConfigs(filePath) {
const metadata = extractAgentMetadata(filePath);
if (!metadata || !metadata.mcpServers) return [];
return Object.entries(metadata.mcpServers).map(([name, cfg]) => {
// Ensure we don't mutate original cfg
const copy = { ...cfg };
return {
name,
type: typeof copy.type === "string" ? copy.type : undefined,
command: typeof copy.command === "string" ? copy.command : undefined,
args: Array.isArray(copy.args) ? copy.args : undefined,
url: typeof copy.url === "string" ? copy.url : undefined,
headers:
typeof copy.headers === "object" && copy.headers !== null
? copy.headers
: undefined,
};
});
}
module.exports = {
parseCollectionYaml,
parseFrontmatter,
extractAgentMetadata,
extractMcpServers,
extractMcpServerConfigs,
safeFileOperation,
};