merge: AgentProject membership — the junction is now load-bearing

Merges feat/agent-project-membership (task 166/152). An agent can finally join
many projects: membership routes, per-membership role and optional brain
override, chat resolving context from the membership, and CLI attach-agent /
detach-agent / get memberships. The legacy Agent.projectId fallback is kept for
this release as the schema comment promises.

Conflict resolved in main.ts: that branch added its RBAC case to
mapUrlToPermission inline, while this branch had extracted the function to
middleware/url-permission.ts so it could be tested. Kept the extraction and
ported the case into it — joining a team is `edit:agents:<name>`, not
`create:agents`, which is the authority to mint a whole agent.

Also fixes something the extraction makes possible. security.test.ts contained a
hand-copied reproduction of mapUrlToPermission, because main.ts self-invokes
main() and nothing in it was importable. A security test asserting against a
copy stops testing the system the moment the real one changes, and this copy had
already drifted. It now imports the real function; all 18 assertions still pass,
against the code that actually runs.

Local: build clean, workspace 2550 passed, lint unchanged at 794 errors.

Co-Authored-By: Claude Opus 5 (1M context) <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01N4wNHWf7xSwnZCWpJcyv9p

.taskmaster/tasks/task_075.md

@@ -0,0 +1,77 @@
+# Task ID: 75
+
+**Title:** Implement section-split Stage
+
+**Status:** done
+
+**Dependencies:** 71 ✓, 74 ✓
+
+**Priority:** high
+
+**Description:** Create the built-in section-split stage that splits content based on detected content type, using structural boundaries for JSON/YAML/XML and headers for prose.
+
+**Details:**
+
+Create `src/mcplocal/src/proxymodel/stages/section-split.ts`:
+
+```typescript
+import type { StageHandler, Section } from '../types';
+import { detectContentType } from '../content-detection';
+
+const handler: StageHandler = async (content, ctx) => {
+  const minSize = (ctx.config.minSectionSize as number) ?? 2000;
+  const maxSize = (ctx.config.maxSectionSize as number) ?? 15000;
+  const contentType = detectContentType(content);
+  
+  let sections: Section[];
+  
+  switch (contentType) {
+    case 'json':
+      sections = splitJson(content, minSize, maxSize);
+      break;
+    case 'yaml':
+      sections = splitYaml(content, minSize, maxSize);
+      break;
+    case 'xml':
+      sections = splitXml(content, minSize, maxSize);
+      break;
+    case 'code':
+      sections = splitCode(content, minSize);
+      break;
+    default:
+      sections = splitProse(content, minSize);
+  }
+  
+  if (sections.length === 0) {
+    return { content, sections: [{ id: 'main', title: 'Content', content }] };
+  }
+  
+  const toc = sections.map((s, i) => `[${s.id}] ${s.title}`).join('\n');
+  return {
+    content: `${sections.length} sections (${contentType}):\n${toc}`,
+    sections,
+  };
+};
+
+function splitJson(content: string, minSize: number, maxSize: number): Section[] {
+  const parsed = JSON.parse(content);
+  if (Array.isArray(parsed)) {
+    return parsed.map((item, i) => ({
+      id: item.id ?? item.name ?? `item-${i}`,
+      title: item.label ?? item.title ?? item.name ?? `Item ${i}`,
+      content: JSON.stringify(item, null, 2),
+    }));
+  }
+  return Object.entries(parsed).map(([key, value]) => ({
+    id: key,
+    title: key,
+    content: JSON.stringify(value, null, 2),
+  }));
+}
+
+// Similar implementations for splitYaml, splitXml, splitCode, splitProse
+```
+
+**Test Strategy:**
+
+Unit tests for each content type: JSON arrays split by element, JSON objects split by key, YAML split by top-level keys, XML split by elements, prose split by markdown headers. Test minSize/maxSize thresholds. Test fallback when content can't be parsed.