diff --git a/.cursor-plugin/marketplace.json b/.cursor-plugin/marketplace.json
index 1f2df9c..b41fc54 100644
--- a/.cursor-plugin/marketplace.json
+++ b/.cursor-plugin/marketplace.json
@@ -12,7 +12,7 @@
     {
       "name": "synter",
       "source": ".",
-      "logo": "https://raw.githubusercontent.com/Synter-Media-AI/plugin/main/assets/logo.png",
+      "logo": "assets/logo.png",
       "description": "The AI Agent Operator for Ads. Connect 21 ad platforms, build audiences, generate on-brand creative, launch cross-platform campaigns, and reallocate spend by ROAS — one conversation, every platform. You direct, the agents execute. Nothing that spends money ships without your approval."
     }
   ]
diff --git a/.cursor-plugin/plugin.json b/.cursor-plugin/plugin.json
index b1e2018..4d926cd 100644
--- a/.cursor-plugin/plugin.json
+++ b/.cursor-plugin/plugin.json
@@ -11,7 +11,7 @@
   "homepage": "https://syntermedia.ai",
   "repository": "https://github.com/Synter-Media-AI/plugin",
   "license": "MIT",
-  "logo": "https://raw.githubusercontent.com/Synter-Media-AI/plugin/main/assets/logo.png",
+  "logo": "assets/logo.png",
   "keywords": [
     "ads",
     "advertising",
@@ -28,7 +28,7 @@
   ],
   "skills": "./skills/",
   "agents": "./agents/",
-  "hooks": "./hooks/cursor-hooks.json",
+  "rules": "./rules/",
   "mcpServers": "./mcp.json",
   "variables": {
     "type": "object",
@@ -36,7 +36,7 @@
       "SYNTER_API_KEY": {
         "type": "string",
         "title": "Synter API key",
-        "description": "Optional for onboarding. Paste your syn_... key from syntermedia.ai/developer to use connected ad accounts. Leave blank to run /synter:quickstart and the free GA4 tools."
+        "description": "Optional for onboarding. Paste your syn_... key from syntermedia.ai/developer to use connected ad accounts. Leave blank to run /synter:quickstart and the GA4 tools."
       }
     }
   }
diff --git a/SECURITY.md b/SECURITY.md
new file mode 100644
index 0000000..d3bef89
--- /dev/null
+++ b/SECURITY.md
@@ -0,0 +1,19 @@
+# Security
+
+## Reporting a vulnerability
+
+Email **joel@synterai.com** with the subject "Security: Synter plugin". Please include steps to reproduce. We'll acknowledge within 2 business days. Please don't open public issues for vulnerabilities.
+
+## What this plugin does
+
+- It contains Markdown skills, agents, and rules, plus one remote MCP server entry. It ships no executable hooks for Cursor.
+- The MCP server is Synter's hosted server over HTTPS. Cursor authenticates with browser OAuth (scopes `tools:read` and `tools:write`). An API key is an optional fallback for headless use and is never stored in this repo.
+
+## Spend and write safety
+
+- Anything that spends money, launches campaigns, or changes budgets requires explicit user approval before it runs. Reads (`list_*`, `get_*`, reports) don't.
+- Actions run only against ad accounts the user has connected and authorized in Synter.
+
+## Data handling
+
+Synter processes ad-account data only to perform the actions you request. See the [Privacy Policy](https://syntermedia.ai/privacy), [Terms](https://syntermedia.ai/terms), and [Security overview](https://syntermedia.ai/security).
diff --git a/hooks/cursor-hooks.json b/hooks/cursor-hooks.json
deleted file mode 100644
index 9b1fa2d..0000000
--- a/hooks/cursor-hooks.json
+++ /dev/null
@@ -1,9 +0,0 @@
-{
-  "hooks": {
-    "sessionStart": [
-      {
-        "command": "./scripts/session-context.sh"
-      }
-    ]
-  }
-}
diff --git a/plugin.json b/plugin.json
deleted file mode 120000
index e22f3da..0000000
--- a/plugin.json
+++ /dev/null
@@ -1 +0,0 @@
-.claude-plugin/plugin.json
\ No newline at end of file
diff --git a/rules/synter-operator.mdc b/rules/synter-operator.mdc
new file mode 100644
index 0000000..00f66c6
--- /dev/null
+++ b/rules/synter-operator.mdc
@@ -0,0 +1,14 @@
+---
+description: Synter operator rules and money-safety guardrails for ad platform work
+alwaysApply: true
+---
+
+You are the Synter operator — the AI Agent Operator for Ads. You direct, the agents execute.
+Operating rules:
+- MONEY SAFETY RULES: Nothing that spends money ships without explicit user approval. No exceptions, no tier carve-outs.
+- Default to recommend-then-execute: propose the plan (platform, budget, audience, creative), show it, and wait for an explicit yes before any create/enable/budget/launch/pause call.
+- Reads run without asking: list_*, get_*, pull_*, verify_*, and reports.
+- Confirm the org/account before any write (list_connected_accounts). Never run one org's campaigns through another's credentials or session. Use real IDs the tools return — never invent them.
+- "Created" != "live and spending" — verify before claiming.
+- Brand voice: terse, certain, doing the work. Show, don't sell. No banned language (no "AI-Powered", "seamless", "beta", hype CTAs). Never fabricate prices, competitors, or stats.
+- Get started or get help: /synter:quickstart and /synter:help.