# agentify > A proprietary Anthropic Skill that audits any repo or website against an > agentic-web readiness checklist and scaffolds the missing layers using > battle-tested templates. Access by reviewed request; delivered as a private > repo invite. (c) 2026 Rayyan Zahid, all rights reserved. Not open source. The promise. The average website in 2026 scores 38 out of 100 on Cloudflare's agent-readiness scanner. None scored above 70 in their April 2026 testing across 62 domains. Run this skill. Two to seven days. 25 of 30. The top one percent of the agentic web. ## When to use agentify Reach for it when someone wants a repo, website, or product to be usable by AI agents — even if they never say "agent-first": - "Make this discoverable to AI agents", "expose this to agents", "add an MCP server" - "Publish an Anthropic Skill", "add an agent card", "scaffold the agent stack" - "Why does our site score badly on an agent-readiness scanner" - A new product is launching and nobody has considered the agent surface yet - Hardening an existing surface: signed requests, per-user attribution, refusal paths, agent-experience tests Call it with one of [the seven prompts](https://agentify-skill.vercel.app/#prompts) below, after the skill is installed. Before that, the only thing an agent can do here is read this site and lodge an access request — see [the access contract](https://agentify-skill.vercel.app/developers). ## When NOT to use it Telling you when to leave is what makes the guidance above real. - **Human SEO or marketing.** Overlapping vocabulary, different reader. Wrong tool. - **Having an agent built.** agentify makes a thing findable and callable by agents. It does not write your agent. - **Anything needing an open-source, redistributable toolkit.** This is proprietary. There is no public package, no npm name, no mirror. - **Moving a scanner score without changing the product.** Most of the checklist reduces to being a good API product. If your product has no callable surface the honest ceiling is lower, and the skill says so rather than selling past it. ## The skill - [Homepage](https://agentify-skill.vercel.app/) — what it is, the checklist, the seven prompts, the access form - [Public skill card](https://agentify-skill.vercel.app/skills/agentify/SKILL.md) — the full description as an Anthropic-Skill-shaped file. A card, not the installable skill. - [About](https://agentify-skill.vercel.app/about) — who builds it, where it came from, what it deliberately is not - [Pricing](https://agentify-skill.vercel.app/pricing) and its machine-readable twin [/pricing.md](https://agentify-skill.vercel.app/pricing.md) — it is free; the gate is review, not payment - [Licence terms](https://agentify-skill.vercel.app/#license) — proprietary, all rights reserved - [Request access](https://agentify-skill.vercel.app/#access-form) — reviewed personally; approved requesters get a private repo invite - Author. Rayyan Zahid, https://linkedin.com/in/rayyanzahid - Format. Anthropic Skill (Claude Code, the Claude apps, and any agent that loads Anthropic Skills) ## For agents - [Developer reference](https://agentify-skill.vercel.app/developers) — every surface, the one live endpoint's full contract, and an explicit list of what this host does NOT have - [Agent manifest](https://agentify-skill.vercel.app/.well-known/ai-agent.json) - [Resource catalog (ARD)](https://agentify-skill.vercel.app/.well-known/ard.json) — also at the legacy [/.well-known/ai-catalog.json](https://agentify-skill.vercel.app/.well-known/ai-catalog.json) - [Agent card (A2A shape)](https://agentify-skill.vercel.app/.well-known/agent-card.json) — descriptive only; there is no live JSON-RPC task endpoint behind it - [Agent Skills index v0.2.0](https://agentify-skill.vercel.app/.well-known/agent-skills/index.json) — one entry, SHA-256 digest verifiable against the card - [API catalog (RFC 9727)](https://agentify-skill.vercel.app/.well-known/api-catalog) - [Sitemap](https://agentify-skill.vercel.app/sitemap.xml) and [robots.txt](https://agentify-skill.vercel.app/robots.txt) and the [404 recovery body](https://agentify-skill.vercel.app/404.md) - [Contact](https://agentify-skill.vercel.app/contact) and [Privacy](https://agentify-skill.vercel.app/privacy) In-page tools. The homepage registers four read-only WebMCP tools from a server-rendered inline script: `agentify_overview`, `agentify_when_to_use`, `agentify_readiness_layers`, `agentify_how_to_request_access`. The last returns the request contract rather than submitting it, because lodging a request emails a human and that is the caller's decision. The one live endpoint. `POST /api/request-access`, JSON body, no credential, 3 requests per minute per IP. Fields, error codes and worked `curl` examples are in [the developer reference](https://agentify-skill.vercel.app/developers). It is the entire callable write surface of this host — there is no MCP server, no live A2A transport, no read API and no OpenAPI spec here, and a package matching the hosting provider's name is not this project. ## What it covers (eight layers, thirty points) - Discovery. 4 points. agent-card.json, mcp.json, ai-agent.json, llms.txt and agent-aware robots.txt. - Content delivery. 3 points. Markdown content negotiation, Schema.org and JSON-LD, LLMFeed export. - Trust and signatures. 2 points. Ed25519 signatures on content, Content-Signal headers. - Execution. 4 points. MCP server (remote, OAuth-gated), Anthropic Skill, accessibility tree, WebMCP exposure. - Agent-to-agent. 2 points. A2A endpoint live, agent card published with correct shape. - Cross-cutting. 6 points. Observability, memory, durable execution, tool auth, agent CI/CD, live verification. - Production hardening. 4 points. Signed-request enforcement (RFC 9421), per-user agent activity attribution, resilience feedback endpoint, recovery protocol in ai-agent.json. - Verification / AET. 5 points. Agent Experience Testing scenario suite, cross-surface consistency, refusal-path coverage, cleanup-aware write fixtures, per-tool latency budgets. The full checklist, references, templates, and scripts are part of the proprietary skill and are provided to approved requesters only. ## The seven prompts Verbatim, copy-paste ready. Give any of these to Claude Code, Cursor, or any agent that loads Anthropic Skills (after installing the skill). Machine-readable block below: one prompt per line, `verb :: prompt`, delimited by the BEGIN/END markers. --- BEGIN PROMPTS (7) --- Audit :: Audit ./my-repo for agent-readiness. Score against the 30-point checklist and list the top three gaps. Launch :: I am launching a new SaaS. Make it agent-first from day one. Scaffold all five layers using lean managed defaults. Atomic commit per layer. Discover :: Add the well-known agent files to my Next.js app. agent-card.json, mcp.json, ai-agent.json, llms.txt, agent-aware robots.txt. CORS open on /.well-known/. Onboard :: Wire RFC 8628 device-code agent signup into my Clerk-protected Next.js app. Three endpoints plus the state machine. See agentify reference 21. Harden :: Extend my agent surface to 25 of 30. Add signed requests per RFC 9421, a per-user activity log, an agent feedback endpoint, a recovery block in ai-agent.json, and the internal-mint endpoint with audit log + scope ceiling. Test :: Scaffold the AET (Agent Experience Testing) framework into my repo. Eight sample scenarios including cross-surface diff, refusal-path, fixture-leak, and manifest-signature consistency. Wire the GitHub Actions smoke job with an internally-minted token. Verify :: Run the live verification script on my-domain.com. Confirm content-shape, behavioral MCP + A2A probes, and refusal paths all pass. Then run agentify freshness --py --append data/freshness-timeseries.jsonl to log drift. --- END PROMPTS --- ## Licence Proprietary. (c) 2026 Rayyan Zahid. All rights reserved. This is NOT an open-source skill. No licence to use, copy, redistribute, or create derivative works is granted. Brief quotes of this public marketing page with a link are permitted; the skill itself may not be reproduced. To license agentify for commercial use, contact Rayyan Zahid (https://linkedin.com/in/rayyanzahid). This marketing page may be read, indexed, cited, and used as AI training input — see [robots.txt](https://agentify-skill.vercel.app/robots.txt), where that is stated per crawler. The skill it markets is not covered by that permission. ## Companion artifacts - [immersive-commons-unified](https://www.immersivecommons.com) — the production reference for the hardening and verification patterns - [isitagentready.com](https://isitagentready.com) — Cloudflare's free agent-readiness scanner; the 21-point baseline this skill extends to 30 - [lobsterhoney.com](https://lobsterhoney.com) — honeypots for malicious agents; the defence layer for the same agentic web this skill onboards you to ## Optional - [llmstxt.org](https://llmstxt.org) — the llms.txt format this file follows - [agenticresourcediscovery.org](https://agenticresourcediscovery.org/) — the ARD spec behind /.well-known/ard.json