feat: support Agent Plugins (#81)

Signed-off-by: Jonah Snider <jonah@jonahsnider.com>

.github/workflows/ci.yml

@@ -23,5 +23,8 @@ jobs:
       - name: Validate skills
         run: npm run validate
 
+      - name: Validate Agent Plugin
+        run: npm run validate:agent-plugin
+
       - name: Validate Cursor plugin
         run: npm run validate:cursor-plugin

CONTRIBUTING.md

@@ -48,7 +48,7 @@ The repo uses npm (not pnpm).
 1. Fork the repo
 2. Install dependencies with `npm ci`
 3. Make your changes in `skills/<skill-name>/`
-4. Run `npm run validate:all` to check skill and plugin validity
+4. Run `npm run validate:all` to check Agent Skills, Agent Plugins, and client plugin validity
 5. Submit a PR
 
 ### Adding to an existing skill

README.md

@@ -13,6 +13,7 @@ Collection of resources to help AI agents build better with [Sanity](https://www
 
 - **MCP server:** Direct access to your Sanity projects (content, datasets, releases, schemas) and agent rules.
 - **Agent skills:** Comprehensive best practices skills for Sanity development, content modeling, SEO/AEO, and experimentation. Includes 21 integration/topic guides and 26 focused best-practice rules.
+- **Agent Plugin:** `plugin.json`, `skills/`, and `mcp.json` components for any Agent Plugins-compatible client.
 - **Claude Code plugin:** MCP server, agent skills, and slash commands for Claude Code users. Available on the [official Anthropic plugin marketplace](https://claude.com/plugins/sanity).
 - **Cursor plugin:** MCP server, agent skills, and commands on the [Cursor Marketplace](https://cursor.com/marketplace/sanity).
 - **Codex plugin:** MCP server and agent skills for [OpenAI Codex](https://developers.openai.com/codex) users.
@@ -25,7 +26,7 @@ Choose your path based on how you want agents to work with Sanity:
 
 1. **MCP server** — Give your agent always up-to-date rules and full access to your Sanity projects. No local files to maintain. Works with Cursor, VS Code, Claude Code, Lovable, v0, Replit, OpenCode, and other MCP-compatible clients.
 2. **Agent skills** — Install best practices skills for Sanity, content modeling, SEO/AEO, and experimentation. Works with Cursor, Claude Code, and any [Agent Skills](https://agentskills.io)-compatible agent.
-3. **Plugin** — Install the Sanity plugin for Cursor or Claude Code. Bundles MCP server, agent skills, and commands.
+3. **Plugin** — Install the Sanity plugin for Cursor, Claude Code, or any Agent Plugins-compatible client. Bundles MCP server, agent skills, and commands.
 4. **Manual installation** — Copy the skill references locally for offline use. You'll need to update them yourself.
 
 ### Option 1: Install MCP server (recommended)
@@ -186,7 +187,8 @@ See [Option 3](#option-3-install-plugin) for plugin installation.
 
 ### Option 3: Install plugin
 
-Install the Sanity plugin to get MCP server, agent skills, and commands. Available on the [Claude Code marketplace](https://claude.com/plugins/sanity) and [Cursor Marketplace](https://cursor.com/marketplace/sanity).
+Install the Sanity plugin to get MCP server, agent skills, and commands.
+Available on the [Claude Code marketplace](https://claude.com/plugins/sanity), [Cursor Marketplace](https://cursor.com/marketplace/sanity), or from this repo as an [Agent Plugin](https://agent-plugins.org/).
 
 #### Claude Code
 
@@ -332,6 +334,8 @@ Just say: "Get started with Sanity" to begin.
 sanity-io/agent-toolkit/
 ├── AGENTS.md                      # Knowledge router & agent behavior
 ├── README.md                      # This file
+├── plugin.json                    # Portable Agent Plugins v1 manifest
+├── mcp.json                       # Portable Agent Plugins v1 MCP configuration
 ├── .agents/plugins/               # Codex marketplace
 │   └── marketplace.json           # Codex marketplace metadata
 ├── .claude-plugin/                # Claude Code plugin configuration (distributed via claude-plugins-official)
@@ -342,7 +346,7 @@ sanity-io/agent-toolkit/
 ├── .cursor-plugin/                # Cursor plugin configuration (distributed via cursor.com/marketplace)
 │   ├── marketplace.json           # Cursor marketplace metadata
 │   └── plugin.json                # Per-plugin manifest
-├── .mcp.json                      # MCP server configuration
+├── .mcp.json                      # Legacy client MCP compatibility configuration
 ├── assets/                        # Plugin branding
 │   └── logo.svg                   # Sanity logo for marketplace display
 ├── commands/                      # Agent commands
@@ -351,6 +355,7 @@ sanity-io/agent-toolkit/
 │   ├── typegen.md                 # /typegen
 │   └── deploy-schema.md           # /deploy-schema
 ├── scripts/                       # Validation and CI scripts
+│   ├── validate-agent-plugin.mjs  # Validate manifests against the v1 JSON Schemas
 │   └── validate-cursor-plugin.mjs # Cursor plugin validator
 └── skills/                        # Agent skills (agentskills.io format)
     ├── sanity-best-practices/     # Comprehensive Sanity skill
@@ -378,6 +383,7 @@ All skills use `references/` for detailed content loaded on demand. The `sanity-
 - [Visual Editing guide](https://www.sanity.io/docs/visual-editing)
 - [Sanity TypeGen](https://www.sanity.io/docs/sanity-typegen)
 - [MCP server docs](https://www.sanity.io/docs/ai/mcp-server)
+- [Agent Plugins specification](https://agent-plugins.org/specification)
 - [Blueprints Infrastructure as Code](https://www.sanity.io/docs/compute-and-ai/blueprints)
 
 ---

mcp.json

@@ -0,0 +1,9 @@
+{
+  "$schema": "https://agent-plugins.org/schemas/1.0.0/mcp.schema.json",
+  "mcpServers": {
+    "Sanity": {
+      "type": "streamable-http",
+      "url": "https://mcp.sanity.io"
+    }
+  }
+}

package-lock.json

@@ -9,7 +9,25 @@
       "version": "1.0.0",
       "license": "MIT",
       "devDependencies": {
-        "skills-ref": "^0.1.5"
+        "ajv": "8.20.0",
+        "skills-ref": "0.1.5"
+      }
+    },
+    "node_modules/ajv": {
+      "version": "8.20.0",
+      "resolved": "https://registry.npmjs.org/ajv/-/ajv-8.20.0.tgz",
+      "integrity": "sha512-Thbli+OlOj+iMPYFBVBfJ3OmCAnaSyNn4M1vz9T6Gka5Jt9ba/HIR56joy65tY6kx/FCF5VXNB819Y7/GUrBGA==",
+      "dev": true,
+      "license": "MIT",
+      "dependencies": {
+        "fast-deep-equal": "^3.1.3",
+        "fast-uri": "^3.0.1",
+        "json-schema-traverse": "^1.0.0",
+        "require-from-string": "^2.0.2"
+      },
+      "funding": {
+        "type": "github",
+        "url": "https://github.com/sponsors/epoberezkin"
       }
     },
     "node_modules/argparse": {
@@ -29,6 +47,30 @@
         "node": ">=18"
       }
     },
+    "node_modules/fast-deep-equal": {
+      "version": "3.1.3",
+      "resolved": "https://registry.npmjs.org/fast-deep-equal/-/fast-deep-equal-3.1.3.tgz",
+      "integrity": "sha512-f3qQ9oQy9j2AhBe/H9VC91wLmKBCCU/gDOnKNAYG5hswO7BLKj09Hc5HYNz9cGI++xlpDCIgDaitVs03ATR84Q==",
+      "dev": true,
+      "license": "MIT"
+    },
+    "node_modules/fast-uri": {
+      "version": "3.1.5",
+      "resolved": "https://registry.npmjs.org/fast-uri/-/fast-uri-3.1.5.tgz",
+      "integrity": "sha512-gHwA1O9LDIcKunMKhObS/HimwtehO1nPUECKAu5TpKgaO19fcWEl4bliWe1jWxVFvIXztJjjQ4L8XQ1EU9f7Jw==",
+      "dev": true,
+      "funding": [
+        {
+          "type": "github",
+          "url": "https://github.com/sponsors/fastify"
+        },
+        {
+          "type": "opencollective",
+          "url": "https://opencollective.com/fastify"
+        }
+      ],
+      "license": "BSD-3-Clause"
+    },
     "node_modules/js-yaml": {
       "version": "4.1.1",
       "resolved": "https://registry.npmjs.org/js-yaml/-/js-yaml-4.1.1.tgz",
@@ -42,6 +84,23 @@
         "js-yaml": "bin/js-yaml.js"
       }
     },
+    "node_modules/json-schema-traverse": {
+      "version": "1.0.0",
+      "resolved": "https://registry.npmjs.org/json-schema-traverse/-/json-schema-traverse-1.0.0.tgz",
+      "integrity": "sha512-NM8/P9n3XjXhIZn1lLhkFaACTOURQXjWhV4BA/RnOv8xvgqtqpAX9IO4mRQxSx1Rlo4tqzeqb0sOlruaOy3dug==",
+      "dev": true,
+      "license": "MIT"
+    },
+    "node_modules/require-from-string": {
+      "version": "2.0.2",
+      "resolved": "https://registry.npmjs.org/require-from-string/-/require-from-string-2.0.2.tgz",
+      "integrity": "sha512-Xf0nWe6RseziFMu+Ap9biiUbmplq6S9/p+7w7YXP/JBHhrUDDUhwa+vANyubuqfZWTveU//DYVGsDG7RKL/vEw==",
+      "dev": true,
+      "license": "MIT",
+      "engines": {
+        "node": ">=0.10.0"
+      }
+    },
     "node_modules/skills-ref": {
       "version": "0.1.5",
       "resolved": "https://registry.npmjs.org/skills-ref/-/skills-ref-0.1.5.tgz",

package.json

@@ -11,16 +11,19 @@
     "validate:experimentation": "skills-ref validate ./skills/content-experimentation-best-practices",
     "validate:pt-serialization": "skills-ref validate ./skills/portable-text-serialization",
     "validate:pt-conversion": "skills-ref validate ./skills/portable-text-conversion",
+    "validate:agent-plugin": "node scripts/validate-agent-plugin.mjs",
     "validate:cursor-plugin": "node scripts/validate-cursor-plugin.mjs",
-    "validate:all": "npm run validate && npm run validate:cursor-plugin"
+    "validate:all": "npm run validate && npm run validate:agent-plugin && npm run validate:cursor-plugin"
   },
   "devDependencies": {
+    "ajv": "8.20.0",
     "skills-ref": "0.1.5"
   },
   "keywords": [
     "sanity",
     "ai",
     "agent",
+    "agent-plugins",
     "skills",
     "claude",
     "cursor",

plugin.json

@@ -0,0 +1,23 @@
+{
+  "$schema": "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
+  "name": "sanity",
+  "version": "1.0.0",
+  "description": "Sanity plugin with an MCP server and agent skills for structured content development.",
+  "author": {
+    "name": "Sanity",
+    "email": "support@sanity.io",
+    "url": "https://www.sanity.io"
+  },
+  "homepage": "https://github.com/sanity-io/agent-toolkit",
+  "repository": "https://github.com/sanity-io/agent-toolkit",
+  "license": "MIT",
+  "keywords": [
+    "sanity",
+    "cms",
+    "content-lake",
+    "groq",
+    "content",
+    "structured-content",
+    "visual-editing"
+  ]
+}

scripts/validate-agent-plugin.mjs

@@ -0,0 +1,44 @@
+#!/usr/bin/env node
+
+import { readFile } from "node:fs/promises";
+import process from "node:process";
+import Ajv2020 from "ajv/dist/2020.js";
+
+const documents = [
+  {
+    path: "plugin.json",
+    schemaUrl:
+      "https://agent-plugins.org/schemas/1.0.0/plugin.schema.json",
+  },
+  {
+    path: "mcp.json",
+    schemaUrl: "https://agent-plugins.org/schemas/1.0.0/mcp.schema.json",
+  },
+];
+
+const ajv = new Ajv2020({ allErrors: true });
+let valid = true;
+
+for (const document of documents) {
+  const [schemaResponse, data] = await Promise.all([
+    fetch(document.schemaUrl),
+    readFile(document.path, "utf8").then(JSON.parse),
+  ]);
+
+  if (!schemaResponse.ok) {
+    throw new Error(
+      `Unable to load ${document.schemaUrl}: ${schemaResponse.status} ${schemaResponse.statusText}`
+    );
+  }
+
+  const validate = ajv.compile(await schemaResponse.json());
+  if (!validate(data)) {
+    valid = false;
+    console.error(`${document.path} is invalid:`);
+    console.error(ajv.errorsText(validate.errors, { separator: "\n" }));
+  }
+}
+
+if (!valid) process.exit(1);
+
+console.log("Agent Plugin manifests are valid.");