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_035.md

@@ -0,0 +1,64 @@
+# Task ID: 35
+
+**Title:** Implement health monitoring across all tiers
+
+**Status:** done
+
+**Dependencies:** 33 ✓, 34 ✓
+
+**Priority:** medium
+
+**Description:** Extend health monitoring to track connectivity and status across mcplocal, mcpd, and individual MCP server instances.
+
+**Details:**
+
+1. Update mcplocal health monitor in `src/mcplocal/src/health.ts`:
+   ```typescript
+   export class TieredHealthMonitor {
+     async checkHealth(): Promise<TieredHealthStatus> {
+       return {
+         mcplocal: {
+           status: 'healthy',
+           llmProvider: await this.checkLlmProvider(),
+           uptime: process.uptime()
+         },
+         mcpd: await this.checkMcpdHealth(),
+         instances: await this.checkInstancesHealth()
+       };
+     }
+     
+     private async checkMcpdHealth(): Promise<McpdHealth> {
+       try {
+         const health = await this.mcpdClient.get('/api/v1/health');
+         return { status: 'connected', ...health };
+       } catch {
+         return { status: 'disconnected' };
+       }
+     }
+     
+     private async checkInstancesHealth(): Promise<InstanceHealth[]> {
+       const instances = await this.mcpdClient.get('/api/v1/instances');
+       return instances.map(i => ({
+         name: i.name,
+         status: i.status,
+         lastHealthCheck: i.lastHealthCheck
+       }));
+     }
+   }
+   ```
+2. Add health endpoint to mcplocal HTTP server: `GET /health`
+3. Update mcpctl status command to display tiered health
+4. Add degraded state detection:
+   - LLM provider unavailable but mcpd reachable
+   - Some instances down but others healthy
+5. Add health event notifications for state transitions
+6. Add configurable health check intervals
+
+**Test Strategy:**
+
+1. Unit test: Health check correctly identifies all states
+2. Unit test: Degraded state is detected correctly
+3. Integration test: Full health check across all tiers
+4. Test health endpoint returns correct format
+5. Test mcpctl status displays health correctly
+6. Test state transition events are emitted