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_052.md

@@ -0,0 +1,63 @@
+# Task ID: 52
+
+**Title:** Implement tool call intercept with briefing injection
+
+**Status:** done
+
+**Dependencies:** 50 ✓, 51 ✓
+
+**Priority:** high
+
+**Description:** When a gated session calls a tool without first calling begin_session, intercept the call, extract keywords, match prompts, and inject the briefing alongside the real tool result.
+
+**Details:**
+
+1. Implement handleInterceptedCall in `/src/mcplocal/src/router.ts`:
+   ```typescript
+   async handleInterceptedCall(
+     sessionId: string,
+     toolName: string,
+     args: any
+   ): Promise<ToolResult> {
+     // 1. Extract keywords from tool call
+     const keywords = this.keywordExtractor.extractKeywords(toolName, args);
+     
+     // 2. Match prompts using keywords
+     const { fullContent, indexOnly, remaining } = 
+       await this.promptSelector.selectPrompts(keywords, this.promptIndex);
+     
+     // 3. Execute the actual tool call
+     const actualResult = await this.routeToolCall(toolName, args);
+     
+     // 4. Build briefing with intercept preamble
+     const briefing = this.buildBriefing(fullContent, indexOnly, remaining, 'intercept');
+     
+     // 5. Ungate session
+     this.ungateSession(sessionId);
+     
+     // 6. Send notifications/tools/list_changed
+     await this.sendToolsListChanged();
+     
+     // 7. Return combined result
+     return {
+       content: [{
+         type: 'text',
+         text: `${briefing}\n\n---\n\n${actualResult.content[0].text}`
+       }]
+     };
+   }
+   ```
+
+2. Use gate-intercept-preamble system prompt for the briefing prefix
+
+**Test Strategy:**
+
+- Unit test: Tool call while gated triggers intercept
+- Unit test: Keywords extracted from tool name and args
+- Unit test: Prompts matched using extracted keywords
+- Unit test: Actual tool still executes and returns result
+- Unit test: Briefing prepended to tool result
+- Unit test: Session ungated after intercept
+- Unit test: notifications/tools/list_changed sent
+- Unit test: Intercept preamble included in briefing
+- Integration test: End-to-end intercept flow