diff --git a/.github/SECURITY.md b/.github/SECURITY.md
new file mode 100644
index 0000000..3b765d5
--- /dev/null
+++ b/.github/SECURITY.md
@@ -0,0 +1,32 @@
+# Security Policy
+
+## Supported Versions
+
+Security updates are provided for the most recent release of `elnora-plugins`. The marketplace and its plugins are versioned together; we recommend always running the latest tag.
+
+| Version | Supported |
+| ------- | --------- |
+| Latest release (`1.x`) | Yes |
+| Older releases | No — please upgrade |
+
+## Reporting a Vulnerability
+
+Please report suspected vulnerabilities privately. Do not open a public issue for security reports.
+
+- **Email:** [security@elnora.ai](mailto:security@elnora.ai)
+- **GitHub:** use [private vulnerability reporting](https://github.com/Elnora-AI/elnora-plugins/security/advisories/new) on this repository
+
+Include the affected version or commit, reproduction steps, and impact. We aim to acknowledge within 3 business days and to provide a remediation timeline after triage. Please give us a reasonable window to release a fix before any public disclosure. We are happy to credit reporters who wish to be named.
+
+## Data Flow, Credentials & Secrets
+
+This repository ships **client-side configuration and Agent Skills only** — it contains no server code and stores no credentials. Understanding what leaves your machine:
+
+- **What it does:** installing a plugin from this marketplace registers the Elnora MCP server (`https://mcp.elnora.ai/mcp`) with your AI tool and copies skill files into that tool's skills directory. When you invoke a skill, your AI client sends requests to the Elnora Platform over HTTPS.
+- **Credentials:** the MCP server authenticates by one of two methods, both handled by your MCP client — never by files in this repo:
+  - **OAuth 2.1** (recommended) — a browser popup completes authorization on first connection; tokens are stored and refreshed by your AI client, not this repository.
+  - **API key** — a Bearer token passed in the `Authorization` header, generated from your Elnora account settings. Keep it in an environment variable or your client's secret store; never commit it or paste it into skill files, configs, or issues.
+- **What leaves your machine:** the prompts, task content, and files you choose to send to the Elnora Platform, plus your auth token, all over TLS to `mcp.elnora.ai`. The skills in this repo do not exfiltrate data to any other host.
+- **No secrets in the repo:** no API keys, tokens, OAuth client secrets, or personal data are committed to this repository. `.env` and `*.local` files are gitignored. If you find a secret committed to git history, treat it as a vulnerability and report it via the channels above.
+
+Because Agent Skills execute in the context of your AI tool, only install marketplaces and skills you trust, and review skill files before running them in a privileged environment.