feat(llm-security-copilot): port llm-security v5.1.0 to GitHub Copilot CLI

Full port of llm-security plugin for internal use on Windows with GitHub
Copilot CLI. Protocol translation layer (copilot-hook-runner.mjs)
normalizes Copilot camelCase I/O to Claude Code snake_case format — all
original hook scripts run unmodified.

- 8 hooks with protocol translation (stdin/stdout/exit code)
- 18 SKILL.md skills (Agent Skills Open Standard)
- 6 .agent.md agent definitions
- 20 scanners + 14 scanner lib modules (unchanged)
- 14 knowledge files (unchanged)
- 39 test files including copilot-port-verify.mjs (17 tests)
- Windows-ready: node:path, os.tmpdir(), process.execPath, no bash

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>

plugins/llm-security-copilot/hooks/scripts/lib/copilot-adapter.mjs

@@ -0,0 +1,73 @@
+#!/usr/bin/env node
+// copilot-adapter.mjs — Normalizes GitHub Copilot hook I/O to internal format.
+//
+// Copilot CLI hooks receive JSON on stdin with camelCase field names
+// (toolName, toolArgs, toolResult) and signal blocking via exit code 2
+// or stdout JSON with permissionDecision: "deny".
+//
+// This adapter provides a consistent interface so hook logic stays
+// platform-agnostic. If Copilot changes field names, only this file
+// needs updating.
+
+import { readFileSync } from 'node:fs';
+
+/**
+ * Parse hook input from stdin. Returns normalized object or null on failure.
+ * Supports both Copilot (camelCase) and Claude Code (snake_case) field names.
+ */
+export function parseInput() {
+  try {
+    const raw = readFileSync(0, 'utf-8');
+    const input = JSON.parse(raw);
+    return {
+      toolName: input.toolName ?? input.tool_name ?? '',
+      toolInput: input.toolArgs ?? input.tool_input ?? {},
+      toolOutput: input.toolResult ?? input.tool_output ?? '',
+      message: input.message ?? {},
+      sessionId: input.sessionId ?? input.session_id ?? '',
+      raw: input,
+    };
+  } catch {
+    return null;
+  }
+}
+
+/**
+ * Block the tool call with a reason message.
+ */
+export function block(reason) {
+  process.stderr.write(reason + '\n');
+  process.stdout.write(JSON.stringify({
+    permissionDecision: 'deny',
+    reason,
+  }));
+  process.exit(2);
+}
+
+/**
+ * Allow the tool call, optionally with an advisory message.
+ */
+export function allow(message) {
+  if (message) {
+    process.stdout.write(JSON.stringify({
+      permissionDecision: 'allow',
+      message,
+    }));
+  }
+  process.exit(0);
+}
+
+/**
+ * Emit a warning to stderr without blocking. Exit 0.
+ */
+export function warn(message) {
+  process.stderr.write(message + '\n');
+  process.exit(0);
+}
+
+/**
+ * Fail-open: exit 0 silently. Used when input can't be parsed.
+ */
+export function failOpen() {
+  process.exit(0);
+}