docs(ecosystem): add OPA MCP (#8618)

Adds **OPA MCP** to the ecosystem listing.

- **Repo:** https://github.com/OrygnsCode/opa-mcp-server
- **npm:** https://www.npmjs.com/package/@orygn/opa-mcp
- **Docker Hub:** https://hub.docker.com/r/orygn/opa-mcp

OPA MCP is a Model Context Protocol server that gives MCP-compatible
clients (Claude Desktop, Cursor, VS Code, Zed, Windsurf, etc.) a
structured interface to Rego. It wraps the OPA CLI, the OPA REST API,
and the Regal linter behind 32 schema-validated tools, plus higher-level
helpers (`rego_explain_decision`, `rego_describe_policy`,
`rego_generate_test_skeleton`, `rego_suggest_fix`) that compose the
primitives into agent-friendly workflows.

## Files

- `docs/src/data/ecosystem/entries/opa-mcp.md` — entry following the
format of `vscode-opa.md` and `regal.md`.
- `docs/static/img/ecosystem-entry-logos/opa-mcp.png` — 256×256 logo,
~55 KB.

## Labels

- `category: tooling`
- `layer: editor`

## docs_features

`learning-rego`, `policy-testing`, `debugging-rego`, `editors` — all
chosen from the existing taxonomy used by other entries.

Signed-off-by: OrygnsCode <38231263+OrygnsCode@users.noreply.github.com>

docs/src/data/ecosystem/entries/opa-mcp.md

@@ -0,0 +1,52 @@
+---
+title: OPA MCP
+subtitle: Model Context Protocol server for authoring and debugging Rego
+labels:
+  category: tooling
+  layer: editor
+inventors:
+- independent
+code:
+- https://github.com/OrygnsCode/opa-mcp-server
+docs_features:
+  learning-rego:
+    note: |
+      OPA MCP exposes higher-level helpers (rego_describe_policy,
+      rego_suggest_fix, rego_explain_decision) that compose opa fmt,
+      opa parse, and opa eval into AI-friendly tools. Agents in any MCP
+      client (Claude Desktop, Cursor, VS Code, Zed) can author, format,
+      and review Rego with structured output instead of free-form CLI text.
+  policy-testing:
+    note: |
+      The rego_test tool runs opa test over a directory and returns
+      pass/fail per test with optional coverage. rego_generate_test_skeleton
+      produces a _test.rego scaffold covering each rule in a policy.
+  debugging-rego:
+    note: |
+      rego_explain_decision wraps opa eval --explain=full and returns a
+      structured trace, letting agents answer "why was this rejected"
+      without reading raw traces. rego_eval_with_profile and
+      rego_eval_with_coverage surface hot rules and per-line coverage.
+  editors:
+    note: |
+      OPA MCP is a stdio MCP server that plugs into any MCP-compatible
+      client. One-line Smithery install for Claude Desktop, drop-in
+      configs for Cursor, VS Code, Zed, and Windsurf. Multi-arch Docker
+      image and a signed .mcpb bundle are also published.
+---
+
+OPA MCP is a Model Context Protocol server that gives MCP-compatible
+clients a structured interface to Rego. It wraps the OPA CLI (opa fmt,
+opa check, opa eval, opa test, opa build, opa sign), the OPA REST API,
+and the Regal linter, exposing 32 tools with stable error codes and
+schema-validated input/output.
+
+Higher-level helpers (rego_explain_decision, rego_describe_policy,
+rego_generate_test_skeleton, rego_suggest_fix) compose the primitives
+into the tasks agents typically perform. A curated resource set exposes
+the OPA built-in function catalog, the Rego style guide, and a pattern
+library covering RBAC, ABAC, Kubernetes admission, IaC gates, API
+authorization, and rate limiting.
+
+Distributed via npm (@orygn/opa-mcp), Docker Hub (orygn/opa-mcp), and a
+signed .mcpb bundle.