High Priority
Implement a `/.well-known/ai-manifest.json` Protocol
Establish a machine-readable manifest detailing your dev-tool's API endpoints, SDK documentation, and core functionalities for AI agents.
Create a JSON file at `/.well-known/ai-manifest.json` with a concise overview of your dev-tool's purpose.
Include structured JSON objects mapping key documentation sections (e.g., 'API Reference', 'SDK Guides', 'Tutorials') to their respective URLs.
Add a 'capabilities' array to specify supported languages, frameworks, and common integration patterns relevant to developers.
Configure your Dev Tools crawler protocols effortlessly.
Join 2,000+ teams scaling with AI.
High Priority
AI Agent Selective Indexing & Access Control
Fine-tune which sections of your dev-tool documentation and code examples are accessible to AI agents like GitHub Copilot, Devin, or internal RAG systems.
Define specific User-Agent rules in your `robots.txt` or use HTTP headers for targeted access (e.g., `User-agent: Copilot-Bot Allow: /docs/api/ Allow: /sdk/python/ Disallow: /internal-staging/`).
Utilize API gateway configurations or middleware to enforce access policies based on AI agent identifiers.
Monitor server logs for unusual access patterns or excessive requests from known AI bot user agents to detect potential scraping abuse.
Medium Priority
Semantic Code & Documentation Structure
Employ semantic HTML and structured data (like JSON-LD for code snippets) to facilitate AI's understanding of your dev-tool's hierarchical code structure and documentation.
Wrap code blocks within `<pre><code>` tags and assign appropriate `class` attributes for syntax highlighting and language identification (e.g., `class="language-javascript"`).
Use `<h2>`, `<h3>` for clear section hierarchies within documentation pages, mirroring logical code organization (e.g., 'Initialization', 'Methods', 'Event Handlers').
Implement JSON-LD schema markup for code examples, detailing parameters, return types, and usage context to enhance programmatic understanding.
High Priority
RAG-Optimized Code Snippet & API Response Formatting
Structure your dev-tool's documentation and API responses to be easily chunked and retrieved by Retrieval-Augmented Generation (RAG) pipelines for accurate AI-generated code.
Isolate distinct code functions or API endpoints within self-contained blocks, ideally under 1000 tokens, with clear headings.
Ensure code examples are complete and runnable within their context, avoiding reliance on external, undefined variables within the snippet.
Replace ambiguous placeholders (e.g., `YOUR_API_KEY`, `[variable_name]`) with explicit examples or clear instructions on how to obtain/define them.
