name: Agent-Ready Docs
description: >-
Capability composition for delivering AI-era documentation. Combines llms.txt
generation, the Fern MCP server, Ask Fern AI search, and Fern's agent-score
audit to make every page consumable by LLMs and coding agents like Claude
Code, Cursor, and Windsurf.
provider: Fern
providerId: fern
created: '2026-05-22'
tags:
- Agent Ready
- AI
- Documentation
- MCP
- llms.txt
workflow:
- step: enable-llms-txt
capability: fern.docs.llmsTxt
description: Turn on llms.txt generation for the docs site.
inputs:
- hostedDocsSite
outputs:
- llmsTxtArtifact
- step: enable-mcp-server
capability: fern.mcp.askFern
description: Register the docs with the Fern MCP server so MCP clients can call ask_fern_ai.
inputs:
- hostedDocsSite
outputs:
- mcpEndpoint
- step: enable-ask-fern
description: Enable Ask Fern in-docs AI search grounded in the customer's docs, SDKs, and guides.
inputs:
- hostedDocsSite
outputs:
- askFernEnabled
- step: audit-agent-readiness
capability: fern.docs.agentScore
description: Run Fern's agent-score tool to quantify how navigable and parseable the docs are for AI agents.
inputs:
- hostedDocsSite
outputs:
- agentReadinessReport
- step: validate-against-clients
description: Run smoke tests against Claude Desktop, Claude Code, Cursor, and Windsurf using the MCP endpoint and llms.txt.
inputs:
- mcpEndpoint
- llmsTxtArtifact
outputs:
- clientCompatibilityReport