@bbearai/mcp-server 0.6.0 → 0.7.0

package/src/index.ts

@@ -145,7 +145,7 @@ const tools = [
         },
         status: {
           type: 'string',
-          enum: ['new', 'triaging', 'confirmed', 'in_progress', 'fixed', 'resolved', 'verified', 'wont_fix', 'duplicate'],
+          enum: ['new', 'triaging', 'confirmed', 'in_progress', 'fixed', 'ready_to_test', 'verified', 'resolved', 'reviewed', 'closed', 'wont_fix', 'duplicate'],
           description: 'The new status for the report',
         },
         resolution_notes: {
@@ -158,7 +158,47 @@ const tools = [
   },
   {
     name: 'get_report_context',
-    description: 'Get the full debugging context for a report including console logs, network requests, and navigation history',
+    description: 'Get the full debugging context for a report including console logs, network requests, and navigation history. Use compact=true for app_context summary only (no console/network/navigation).',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        report_id: {
+          type: 'string',
+          description: 'The UUID of the report',
+        },
+        compact: {
+          type: 'boolean',
+          description: 'Compact mode: returns app_context only, skips console logs, network requests, and navigation history. (default: false)',
+        },
+      },
+      required: ['report_id'],
+    },
+  },
+  {
+    name: 'add_report_comment',
+    description: 'Add a comment/note to a bug report thread without changing its status. Use this for follow-up questions, investigation notes, or developer-tester communication.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        report_id: {
+          type: 'string',
+          description: 'The UUID of the report to comment on',
+        },
+        message: {
+          type: 'string',
+          description: 'The comment/note content',
+        },
+        author: {
+          type: 'string',
+          description: 'Optional author name (defaults to "Claude Code")',
+        },
+      },
+      required: ['report_id', 'message'],
+    },
+  },
+  {
+    name: 'get_report_comments',
+    description: 'Get all comments/notes on a bug report in chronological order. Returns the full discussion thread.',
     inputSchema: {
       type: 'object' as const,
       properties: {
@@ -325,7 +365,7 @@ const tools = [
   },
   {
     name: 'list_test_cases',
-    description: 'List all test cases in the project. Returns test_key, title, target_route, and other metadata. Use this to see existing tests before updating them.',
+    description: 'List all test cases in the project. Returns test_key, title, target_route, and other metadata. Use this to see existing tests before updating them. Use compact=true for id, test_key, title, and priority only (saves tokens).',
     inputSchema: {
       type: 'object' as const,
       properties: {
@@ -350,6 +390,10 @@ const tools = [
           type: 'number',
           description: 'Offset for pagination (default 0)',
         },
+        compact: {
+          type: 'boolean',
+          description: 'Compact mode: returns id, test_key, title, and priority only. (default: false)',
+        },
       },
     },
   },
@@ -458,7 +502,7 @@ const tools = [
         },
         notify_tester: {
           type: 'boolean',
-          description: 'If true, notify the original tester about the fix with a message and verification task. Default: false (silent resolve).',
+          description: 'Notify the original tester about the fix with a message and verification task. Default: true. Set to false for silent resolve.',
         },
       },
       required: ['report_id', 'commit_sha'],
@@ -700,17 +744,17 @@ const tools = [
   },
   {
     name: 'get_coverage_matrix',
-    description: 'Get a comprehensive Route × Track coverage matrix showing test counts, pass rates, and execution data. Use this for a complete view of test coverage.',
+    description: 'Get a comprehensive Route × Track coverage matrix showing test counts, pass rates, and execution data. Use this for a complete view of test coverage. Execution data and bug counts are opt-in to save tokens.',
     inputSchema: {
       type: 'object' as const,
       properties: {
         include_execution_data: {
           type: 'boolean',
-          description: 'Include pass/fail rates and last execution times (default: true)',
+          description: 'Include pass/fail rates and last execution times (default: false). Set true when you need execution history.',
         },
         include_bug_counts: {
           type: 'boolean',
-          description: 'Include open/critical bug counts per route (default: true)',
+          description: 'Include open/critical bug counts per route (default: false). Set true when you need bug context.',
         },
       },
     },
@@ -944,6 +988,11 @@ const tools = [
           enum: ['ios', 'android', 'web'],
           description: 'Filter by platform support',
         },
+        role: {
+          type: 'string',
+          enum: ['tester', 'feedback'],
+          description: 'Filter by role: "tester" for QA testers, "feedback" for feedback-only users (default: all)',
+        },
       },
     },
   },
@@ -1032,6 +1081,21 @@ const tools = [
       required: ['tester_id', 'test_case_ids'],
     },
   },
+  {
+    name: 'unassign_tests',
+    description: 'Remove one or more test assignments by assignment ID. Preserves the test case and its history — only the assignment link is deleted. Use list_test_assignments first to find assignment IDs. Max 50 per call.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        assignment_ids: {
+          type: 'array',
+          items: { type: 'string' },
+          description: 'Array of test assignment UUIDs to remove (required, max 50)',
+        },
+      },
+      required: ['assignment_ids'],
+    },
+  },
   {
     name: 'get_tester_workload',
     description: 'View a specific tester\'s current workload — assignment counts by status and recent assignments.',
@@ -1074,6 +1138,11 @@ const tools = [
           type: 'string',
           description: 'Optional notes about the tester',
         },
+        role: {
+          type: 'string',
+          enum: ['tester', 'feedback'],
+          description: 'Role: "tester" for QA testers (default), "feedback" for feedback-only users',
+        },
       },
       required: ['name', 'email'],
     },
@@ -1176,7 +1245,7 @@ const tools = [
   },
   {
     name: 'export_test_results',
-    description: 'Export test results for a specific test run as structured JSON — includes every assignment, tester, result, and duration.',
+    description: 'Export test results for a specific test run as structured JSON — includes every assignment, tester, result, and duration. Use compact=true for summary only (no assignments array). Use limit to cap assignments returned.',
     inputSchema: {
       type: 'object' as const,
       properties: {
@@ -1184,6 +1253,14 @@ const tools = [
           type: 'string',
           description: 'UUID of the test run to export (required)',
         },
+        compact: {
+          type: 'boolean',
+          description: 'Compact mode: returns test run info + summary only, no assignments array. (default: false)',
+        },
+        limit: {
+          type: 'number',
+          description: 'Max assignments to return in full mode (default: 100, max: 500). Ignored when compact=true.',
+        },
       },
       required: ['test_run_id'],
     },
@@ -1232,8 +1309,677 @@ const tools = [
       properties: {},
     },
   },
+  // === TEST EXECUTION INTELLIGENCE ===
+  {
+    name: 'get_test_impact',
+    description: 'Given changed files, identify which test cases are affected by mapping file paths to test case target routes.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        changed_files: {
+          type: 'array',
+          items: { type: 'string' },
+          description: 'List of changed file paths (relative to project root)',
+        },
+      },
+      required: ['changed_files'],
+    },
+  },
+  {
+    name: 'get_flaky_tests',
+    description: 'Analyze test run history to identify tests with intermittent failure rates above a threshold.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        threshold: {
+          type: 'number',
+          description: 'Minimum flakiness rate to report (0-100, default: 5)',
+        },
+        limit: {
+          type: 'number',
+          description: 'Maximum results to return (default: 20)',
+        },
+      },
+    },
+  },
+  {
+    name: 'assess_test_quality',
+    description: 'Analyze test case steps for weak patterns: vague assertions, missing edge cases, no negative testing, generic descriptions.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        test_case_ids: {
+          type: 'array',
+          items: { type: 'string' },
+          description: 'Specific test case IDs to assess. If omitted, assesses recent test cases.',
+        },
+        limit: {
+          type: 'number',
+          description: 'Maximum test cases to assess (default: 20)',
+        },
+      },
+    },
+  },
+  {
+    name: 'get_test_execution_summary',
+    description: 'Aggregate test execution metrics: pass rate, completion rate, most-failed tests, fastest/slowest tests.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        days: {
+          type: 'number',
+          description: 'Number of days to analyze (default: 30)',
+        },
+      },
+    },
+  },
+  {
+    name: 'check_test_freshness',
+    description: 'Identify test cases that have not been updated since their target code was modified.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        limit: {
+          type: 'number',
+          description: 'Maximum results to return (default: 20)',
+        },
+      },
+    },
+  },
+  {
+    name: 'get_untested_changes',
+    description: 'Given recent commits or changed files, find code changes with no corresponding test coverage in BugBear.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        changed_files: {
+          type: 'array',
+          items: { type: 'string' },
+          description: 'List of changed file paths. If omitted, uses git diff against main.',
+        },
+      },
+    },
+  },
+  // === AUTO-MONITORING TOOLS ===
+  {
+    name: 'get_auto_detected_issues',
+    description: 'Get auto-detected monitoring issues grouped by error fingerprint. Shows recurring crashes, API failures, and rage clicks with frequency and user impact.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        source: {
+          type: 'string',
+          enum: ['auto_crash', 'auto_api', 'auto_rage_click'],
+          description: 'Filter by source type',
+        },
+        min_occurrences: {
+          type: 'number',
+          description: 'Min occurrence count (default: 1)',
+        },
+        since: {
+          type: 'string',
+          description: 'ISO date — only issues after this date (default: 7 days ago)',
+        },
+        limit: {
+          type: 'number',
+          description: 'Max results (default: 20)',
+        },
+        compact: {
+          type: 'boolean',
+          description: 'Compact mode: fingerprint, source, count only',
+        },
+      },
+    },
+  },
+  {
+    name: 'generate_tests_from_errors',
+    description: 'Suggest QA test cases from auto-detected error patterns. Returns structured suggestions — does NOT auto-create test cases.',
+    inputSchema: {
+      type: 'object' as const,
+      properties: {
+        report_ids: {
+          type: 'array',
+          items: { type: 'string' },
+          description: 'Specific report IDs. If omitted, uses top uncovered errors.',
+        },
+        limit: {
+          type: 'number',
+          description: 'Max suggestions (default: 5)',
+        },
+      },
+    },
+  },
 ];
 
+// === TEST EXECUTION INTELLIGENCE ===
+
+async function getTestImpact(args: { changed_files: string[] }) {
+  const projectId = requireProject();
+  const changedFiles = args.changed_files || [];
+
+  if (changedFiles.length === 0) {
+    return { affectedTests: [], message: 'No changed files provided.' };
+  }
+
+  // Get all test cases for the project with their target routes
+  const { data: testCases, error } = await supabase
+    .from('test_cases')
+    .select('id, title, target_route, qa_track, priority')
+    .eq('project_id', projectId);
+
+  if (error) return { error: error.message };
+  if (!testCases || testCases.length === 0) {
+    return { affectedTests: [], message: 'No test cases found for this project.' };
+  }
+
+  // Map changed files to affected test cases
+  const affected: Array<{ testId: string; title: string; targetRoute: string; matchedFiles: string[]; qaTrack: string }> = [];
+
+  for (const tc of testCases) {
+    const route = tc.target_route || '';
+    const matchedFiles = changedFiles.filter(f => {
+      // Match file path to route (e.g., src/app/api/tasks/route.ts -> /api/tasks)
+      const normalized = f.replace(/\\/g, '/');
+      const routeParts = route.split('/').filter(Boolean);
+      return routeParts.some((part: string) => normalized.includes(part)) || normalized.includes(route.replace(/\//g, '/'));
+    });
+
+    if (matchedFiles.length > 0) {
+      affected.push({
+        testId: tc.id,
+        title: tc.title,
+        targetRoute: route,
+        matchedFiles,
+        qaTrack: tc.qa_track,
+      });
+    }
+  }
+
+  return {
+    affectedTests: affected,
+    totalTestCases: testCases.length,
+    affectedCount: affected.length,
+    changedFileCount: changedFiles.length,
+  };
+}
+
+async function getFlakyTests(args: { threshold?: number; limit?: number }) {
+  const projectId = requireProject();
+  const threshold = args.threshold || 5;
+  const limit = args.limit || 20;
+
+  // Get test results grouped by test case
+  const { data: results, error } = await supabase
+    .from('test_results')
+    .select('test_case_id, status, test_cases!inner(title, target_route, qa_track)')
+    .eq('test_cases.project_id', projectId)
+    .order('created_at', { ascending: false })
+    .limit(5000);
+
+  if (error) return { error: error.message };
+  if (!results || results.length === 0) {
+    return { flakyTests: [], message: 'No test results found.' };
+  }
+
+  // Group by test case and calculate flakiness
+  const testStats: Record<string, { passes: number; fails: number; total: number; title: string; route: string; track: string }> = {};
+
+  for (const r of results) {
+    const id = r.test_case_id;
+    if (!testStats[id]) {
+      const tc = r.test_cases as any;
+      testStats[id] = { passes: 0, fails: 0, total: 0, title: tc?.title || '', route: tc?.target_route || '', track: tc?.qa_track || '' };
+    }
+    testStats[id].total++;
+    if (r.status === 'pass') testStats[id].passes++;
+    else if (r.status === 'fail') testStats[id].fails++;
+  }
+
+  // Find flaky tests (have both passes and fails, with fail rate above threshold)
+  const flaky = Object.entries(testStats)
+    .filter(([, stats]) => {
+      if (stats.total < 3) return false; // Need enough data
+      const failRate = (stats.fails / stats.total) * 100;
+      const passRate = (stats.passes / stats.total) * 100;
+      return failRate >= threshold && passRate > 0; // Has both passes and fails
+    })
+    .map(([id, stats]) => ({
+      testCaseId: id,
+      title: stats.title,
+      targetRoute: stats.route,
+      qaTrack: stats.track,
+      totalRuns: stats.total,
+      failRate: Math.round((stats.fails / stats.total) * 100),
+      passRate: Math.round((stats.passes / stats.total) * 100),
+    }))
+    .sort((a, b) => b.failRate - a.failRate)
+    .slice(0, limit);
+
+  return {
+    flakyTests: flaky,
+    totalAnalyzed: Object.keys(testStats).length,
+    flakyCount: flaky.length,
+    threshold,
+  };
+}
+
+async function assessTestQuality(args: { test_case_ids?: string[]; limit?: number }) {
+  const projectId = requireProject();
+  const limit = args.limit || 20;
+
+  let query = supabase
+    .from('test_cases')
+    .select('id, title, steps, target_route, qa_track, priority')
+    .eq('project_id', projectId)
+    .limit(limit);
+
+  if (args.test_case_ids && args.test_case_ids.length > 0) {
+    query = query.in('id', args.test_case_ids);
+  }
+
+  const { data: testCases, error } = await query;
+  if (error) return { error: error.message };
+  if (!testCases || testCases.length === 0) {
+    return { assessments: [], message: 'No test cases found.' };
+  }
+
+  const assessments = testCases.map(tc => {
+    const issues: string[] = [];
+    const steps = tc.steps || [];
+
+    // Check for weak patterns
+    if (steps.length < 2) {
+      issues.push('Too few steps — test may not cover the full flow');
+    }
+
+    const allStepsText = steps.map((s: any) => (typeof s === 'string' ? s : s.action || s.description || '')).join(' ');
+
+    // Vague assertions
+    if (/should work|looks good|is correct|verify it works/i.test(allStepsText)) {
+      issues.push('Vague assertions detected — use specific expected outcomes');
+    }
+
+    // Missing edge cases
+    if (!/error|invalid|empty|missing|unauthorized|forbidden|404|500/i.test(allStepsText)) {
+      issues.push('No negative/error test cases — add edge case testing');
+    }
+
+    // Generic descriptions
+    if (/test the|check the|verify the/i.test(tc.title) && tc.title.length < 30) {
+      issues.push('Generic test title — be more specific about what is being tested');
+    }
+
+    // No specific UI elements referenced
+    if (!/button|input|form|modal|dropdown|select|click|type|enter|submit/i.test(allStepsText)) {
+      issues.push('No specific UI elements referenced — steps may be too abstract');
+    }
+
+    const quality = issues.length === 0 ? 'good' : issues.length <= 2 ? 'needs-improvement' : 'poor';
+
+    return {
+      testCaseId: tc.id,
+      title: tc.title,
+      targetRoute: tc.target_route,
+      stepCount: steps.length,
+      quality,
+      issues,
+    };
+  });
+
+  const qualityCounts = {
+    good: assessments.filter(a => a.quality === 'good').length,
+    needsImprovement: assessments.filter(a => a.quality === 'needs-improvement').length,
+    poor: assessments.filter(a => a.quality === 'poor').length,
+  };
+
+  return {
+    assessments,
+    summary: qualityCounts,
+    totalAssessed: assessments.length,
+  };
+}
+
+async function getTestExecutionSummary(args: { days?: number }) {
+  const projectId = requireProject();
+  const days = args.days || 30;
+  const since = new Date(Date.now() - days * 24 * 60 * 60 * 1000).toISOString();
+
+  // Get test results
+  const { data: results, error } = await supabase
+    .from('test_results')
+    .select('test_case_id, status, duration_ms, created_at, test_cases!inner(title, target_route)')
+    .eq('test_cases.project_id', projectId)
+    .gte('created_at', since)
+    .order('created_at', { ascending: false });
+
+  if (error) return { error: error.message };
+  if (!results || results.length === 0) {
+    return { message: `No test results found in the last ${days} days.` };
+  }
+
+  const totalRuns = results.length;
+  const passed = results.filter(r => r.status === 'pass').length;
+  const failed = results.filter(r => r.status === 'fail').length;
+  const blocked = results.filter(r => r.status === 'blocked').length;
+
+  // Most failed tests
+  const failCounts: Record<string, { count: number; title: string; route: string }> = {};
+  for (const r of results.filter(r => r.status === 'fail')) {
+    const id = r.test_case_id;
+    const tc = r.test_cases as any;
+    if (!failCounts[id]) {
+      failCounts[id] = { count: 0, title: tc?.title || '', route: tc?.target_route || '' };
+    }
+    failCounts[id].count++;
+  }
+  const mostFailed = Object.entries(failCounts)
+    .sort((a, b) => b[1].count - a[1].count)
+    .slice(0, 5)
+    .map(([id, data]) => ({ testCaseId: id, ...data }));
+
+  // Duration stats
+  const durations = results.filter(r => r.duration_ms).map(r => r.duration_ms as number);
+  const avgDuration = durations.length > 0 ? Math.round(durations.reduce((a, b) => a + b, 0) / durations.length) : 0;
+  const maxDuration = durations.length > 0 ? Math.max(...durations) : 0;
+
+  return {
+    period: `${days} days`,
+    totalRuns,
+    passRate: Math.round((passed / totalRuns) * 100),
+    failRate: Math.round((failed / totalRuns) * 100),
+    blockedCount: blocked,
+    averageDurationMs: avgDuration,
+    maxDurationMs: maxDuration,
+    mostFailed,
+    uniqueTestsCovered: new Set(results.map(r => r.test_case_id)).size,
+  };
+}
+
+async function checkTestFreshness(args: { limit?: number }) {
+  const projectId = requireProject();
+  const limit = args.limit || 20;
+
+  // Get test cases with their last update and last result
+  const { data: testCases, error } = await supabase
+    .from('test_cases')
+    .select('id, title, target_route, updated_at, created_at')
+    .eq('project_id', projectId)
+    .order('updated_at', { ascending: true })
+    .limit(limit);
+
+  if (error) return { error: error.message };
+  if (!testCases || testCases.length === 0) {
+    return { staleTests: [], message: 'No test cases found.' };
+  }
+
+  const thirtyDaysAgo = new Date(Date.now() - 30 * 24 * 60 * 60 * 1000).toISOString();
+
+  const stale = testCases
+    .filter(tc => tc.updated_at < thirtyDaysAgo)
+    .map(tc => ({
+      testCaseId: tc.id,
+      title: tc.title,
+      targetRoute: tc.target_route,
+      lastUpdated: tc.updated_at,
+      daysSinceUpdate: Math.round((Date.now() - new Date(tc.updated_at).getTime()) / (24 * 60 * 60 * 1000)),
+    }));
+
+  return {
+    staleTests: stale,
+    totalTestCases: testCases.length,
+    staleCount: stale.length,
+    stalenessThreshold: '30 days',
+  };
+}
+
+async function getUntestedChanges(args: { changed_files?: string[] }) {
+  const projectId = requireProject();
+
+  // Get all test cases to understand what's covered
+  const { data: testCases, error } = await supabase
+    .from('test_cases')
+    .select('id, title, target_route')
+    .eq('project_id', projectId);
+
+  if (error) return { error: error.message };
+
+  const coveredRoutes = new Set((testCases || []).map(tc => tc.target_route).filter(Boolean));
+
+  // If changed_files provided, check coverage
+  const changedFiles = args.changed_files || [];
+
+  if (changedFiles.length === 0) {
+    return {
+      message: 'No changed files provided. Pass changed_files to check coverage.',
+      totalCoveredRoutes: coveredRoutes.size,
+    };
+  }
+
+  // Map changed files to routes and check coverage
+  const untested: Array<{ file: string; inferredRoute: string; reason: string }> = [];
+
+  for (const file of changedFiles) {
+    const normalized = file.replace(/\\/g, '/');
+
+    // Extract route-like path from file
+    let inferredRoute = '';
+
+    // Next.js app router: app/api/tasks/route.ts -> /api/tasks
+    const appRouterMatch = normalized.match(/app\/(api\/[^/]+(?:\/[^/]+)*?)\/route\.\w+$/);
+    if (appRouterMatch) {
+      inferredRoute = '/' + appRouterMatch[1];
+    }
+
+    // Pages router: pages/api/tasks.ts -> /api/tasks
+    const pagesMatch = normalized.match(/pages\/(api\/[^.]+)\.\w+$/);
+    if (!inferredRoute && pagesMatch) {
+      inferredRoute = '/' + pagesMatch[1];
+    }
+
+    // Component files
+    const componentMatch = normalized.match(/(?:components|screens|pages)\/([^.]+)\.\w+$/);
+    if (!inferredRoute && componentMatch) {
+      inferredRoute = '/' + componentMatch[1].replace(/\\/g, '/');
+    }
+
+    if (inferredRoute && !coveredRoutes.has(inferredRoute)) {
+      untested.push({
+        file,
+        inferredRoute,
+        reason: 'No test cases cover this route',
+      });
+    } else if (!inferredRoute) {
+      // Can't map to a route — flag as potentially untested
+      untested.push({
+        file,
+        inferredRoute: 'unknown',
+        reason: 'Could not map file to a testable route',
+      });
+    }
+  }
+
+  return {
+    untestedChanges: untested,
+    changedFileCount: changedFiles.length,
+    untestedCount: untested.length,
+    coveredRoutes: coveredRoutes.size,
+  };
+}
+
+// === AUTO-MONITORING HANDLERS ===
+
+async function getAutoDetectedIssues(args: {
+  source?: string;
+  min_occurrences?: number;
+  since?: string;
+  limit?: number;
+  compact?: boolean;
+}) {
+  const projectId = requireProject();
+  const since = args.since || new Date(Date.now() - 7 * 24 * 60 * 60 * 1000).toISOString();
+  const limit = args.limit || 20;
+
+  let query = supabase
+    .from('reports')
+    .select('id, error_fingerprint, report_source, title, severity, reporter_id, sentry_event_id, created_at, app_context')
+    .eq('project_id', projectId)
+    .neq('report_source', 'manual')
+    .not('error_fingerprint', 'is', null)
+    .gte('created_at', since)
+    .order('created_at', { ascending: false });
+
+  if (args.source) {
+    query = query.eq('report_source', args.source);
+  }
+
+  const { data, error } = await query;
+  if (error) return { error: error.message };
+  if (!data || data.length === 0) return { issues: [], total: 0 };
+
+  // Group by fingerprint
+  const grouped = new Map<string, typeof data>();
+  for (const report of data) {
+    const fp = report.error_fingerprint!;
+    if (!grouped.has(fp)) grouped.set(fp, []);
+    grouped.get(fp)!.push(report);
+  }
+
+  // Build issue summaries
+  const issues = Array.from(grouped.entries())
+    .map(([fingerprint, reports]) => {
+      const uniqueReporters = new Set(reports.map(r => r.reporter_id));
+      const sorted = reports.sort((a, b) => new Date(a.created_at).getTime() - new Date(b.created_at).getTime());
+      const first = sorted[0];
+      const last = sorted[sorted.length - 1];
+      const route = (first.app_context as any)?.currentRoute || 'unknown';
+
+      return {
+        fingerprint,
+        source: first.report_source,
+        message: first.title,
+        route,
+        occurrence_count: reports.length,
+        affected_users: uniqueReporters.size,
+        first_seen: first.created_at,
+        last_seen: last.created_at,
+        severity: first.severity,
+        has_sentry_link: reports.some(r => r.sentry_event_id != null),
+        sample_report_id: first.id,
+      };
+    })
+    .filter(issue => issue.occurrence_count >= (args.min_occurrences || 1))
+    .sort((a, b) => b.occurrence_count - a.occurrence_count)
+    .slice(0, limit);
+
+  if (args.compact) {
+    return {
+      issues: issues.map(i => ({
+        fingerprint: i.fingerprint,
+        source: i.source,
+        count: i.occurrence_count,
+        users: i.affected_users,
+        severity: i.severity,
+      })),
+      total: issues.length,
+    };
+  }
+
+  return { issues, total: issues.length };
+}
+
+async function generateTestsFromErrors(args: {
+  report_ids?: string[];
+  limit?: number;
+}) {
+  const projectId = requireProject();
+  const limit = args.limit || 5;
+
+  let reports;
+  if (args.report_ids?.length) {
+    // Validate all UUIDs
+    for (const id of args.report_ids) {
+      if (!isValidUUID(id)) {
+        return { error: `Invalid report_id format: ${id}` };
+      }
+    }
+
+    const { data, error } = await supabase
+      .from('reports')
+      .select('id, title, report_source, severity, app_context, error_fingerprint, description')
+      .eq('project_id', projectId)
+      .in('id', args.report_ids);
+    if (error) return { error: error.message };
+    reports = data;
+  } else {
+    // Get top uncovered auto-detected errors
+    const { data, error } = await supabase
+      .from('reports')
+      .select('id, title, report_source, severity, app_context, error_fingerprint, description')
+      .eq('project_id', projectId)
+      .neq('report_source', 'manual')
+      .not('error_fingerprint', 'is', null)
+      .order('created_at', { ascending: false })
+      .limit(50);
+    if (error) return { error: error.message };
+
+    // Deduplicate by fingerprint, keep first occurrence
+    const seen = new Set<string>();
+    reports = (data || []).filter(r => {
+      if (!r.error_fingerprint || seen.has(r.error_fingerprint)) return false;
+      seen.add(r.error_fingerprint);
+      return true;
+    }).slice(0, limit);
+  }
+
+  if (!reports?.length) return { suggestions: [] };
+
+  const suggestions = reports.map(report => {
+    const route = (report.app_context as any)?.currentRoute || '/unknown';
+    const source = report.report_source;
+    const priority = report.severity === 'critical' ? 'P1' : report.severity === 'high' ? 'P1' : 'P2';
+
+    let suggestedSteps: string[];
+    if (source === 'auto_crash') {
+      suggestedSteps = [
+        `Navigate to ${route}`,
+        'Reproduce the action that triggered the crash',
+        'Verify the page does not throw an unhandled error',
+        'Verify error boundary displays a user-friendly message if error occurs',
+      ];
+    } else if (source === 'auto_api') {
+      const statusCode = (report.app_context as any)?.custom?.statusCode || 'error';
+      const method = (report.app_context as any)?.custom?.requestMethod || 'API';
+      suggestedSteps = [
+        `Navigate to ${route}`,
+        `Trigger the ${method} request that returned ${statusCode}`,
+        'Verify the request succeeds or displays an appropriate error message',
+        'Verify no data corruption occurs on failure',
+      ];
+    } else {
+      // rage_click or sentry_sync
+      const target = (report.app_context as any)?.custom?.targetSelector || 'the element';
+      suggestedSteps = [
+        `Navigate to ${route}`,
+        `Click on ${target}`,
+        'Verify the element responds to interaction',
+        'Verify loading state is shown if action takes time',
+      ];
+    }
+
+    return {
+      title: `Test: ${report.title?.replace('[Auto] ', '') || 'Auto-detected issue'}`,
+      track: source === 'auto_crash' ? 'Stability' : source === 'auto_api' ? 'API' : 'UX',
+      priority,
+      rationale: `Auto-detected ${source?.replace('auto_', '')} on ${route}`,
+      suggested_steps: suggestedSteps,
+      source_report_id: report.id,
+      route,
+    };
+  });
+
+  return { suggestions };
+}
+
 // === Project management handlers ===
 
 async function listProjects() {
@@ -1358,7 +2104,7 @@ async function getReport(args: { report_id: string }) {
       app_context: data.app_context,
       device_info: data.device_info,
       navigation_history: data.navigation_history,
-      screenshots: data.screenshots,
+      screenshot_urls: data.screenshot_urls,
       created_at: data.created_at,
       reporter: data.tester ? {
         name: data.tester.name,
@@ -1444,14 +2190,14 @@ async function updateReportStatus(args: {
   return { success: true, message: `Report status updated to ${args.status}` };
 }
 
-async function getReportContext(args: { report_id: string }) {
+async function getReportContext(args: { report_id: string; compact?: boolean }) {
   if (!isValidUUID(args.report_id)) {
     return { error: 'Invalid report_id format' };
   }
 
   const { data, error } = await supabase
     .from('reports')
-    .select('app_context, device_info, navigation_history, enhanced_context')
+    .select('app_context, device_info, navigation_history, enhanced_context, screenshot_urls')
     .eq('id', args.report_id)
     .eq('project_id', currentProjectId) // Security: ensure report belongs to this project
     .single();
@@ -1460,16 +2206,83 @@ async function getReportContext(args: { report_id: string }) {
     return { error: error.message };
   }
 
+  // Compact: return app_context only (skip console/network/navigation)
+  if (args.compact === true) {
+    return {
+      context: {
+        app_context: data.app_context,
+        screenshot_urls: data.screenshot_urls,
+      },
+    };
+  }
+
   return {
     context: {
       app_context: data.app_context,
       device_info: data.device_info,
       navigation_history: data.navigation_history,
       enhanced_context: data.enhanced_context || {},
+      screenshot_urls: data.screenshot_urls,
     },
   };
 }
 
+async function addReportComment(args: { report_id: string; message: string; author?: string }) {
+  if (!isValidUUID(args.report_id)) return { error: 'Invalid report_id format' };
+  if (!args.message?.trim()) return { error: 'Message is required' };
+
+  // Verify report exists
+  const { data: report } = await supabase
+    .from('reports').select('id').eq('id', args.report_id).eq('project_id', currentProjectId).single();
+  if (!report) return { error: 'Report not found' };
+
+  // Find or create a discussion thread for this report
+  const { data: existingThread } = await supabase
+    .from('discussion_threads').select('id')
+    .eq('project_id', currentProjectId).eq('report_id', args.report_id).eq('thread_type', 'report')
+    .limit(1).single();
+
+  let threadId: string;
+  if (existingThread) {
+    threadId = existingThread.id;
+  } else {
+    const newId = crypto.randomUUID();
+    const { error: threadErr } = await supabase
+      .from('discussion_threads').insert({
+        id: newId, project_id: currentProjectId, report_id: args.report_id,
+        thread_type: 'report', subject: 'Bug Report Discussion', audience: 'all',
+        priority: 'normal', created_by_admin: true, last_message_at: new Date().toISOString(),
+      });
+    if (threadErr) return { error: `Failed to create thread: ${threadErr.message}` };
+    threadId = newId;
+  }
+
+  const { data: msg, error: msgErr } = await supabase
+    .from('discussion_messages').insert({
+      thread_id: threadId, sender_type: 'admin', sender_name: args.author || 'Claude Code', content: args.message.trim(), content_type: 'text',
+    }).select('id, content, created_at').single();
+
+  if (msgErr) return { error: `Failed to add comment: ${msgErr.message}` };
+  return { success: true, comment: { id: msg.id, thread_id: threadId, content: msg.content, author: args.author || 'Claude Code', created_at: msg.created_at }, message: 'Comment added to report' };
+}
+
+async function getReportComments(args: { report_id: string }) {
+  if (!isValidUUID(args.report_id)) return { error: 'Invalid report_id format' };
+
+  const { data: threads } = await supabase
+    .from('discussion_threads').select('id')
+    .eq('project_id', currentProjectId).eq('report_id', args.report_id).order('created_at', { ascending: true });
+
+  if (!threads || threads.length === 0) return { comments: [], total: 0, message: 'No comments on this report' };
+
+  const { data: messages, error } = await supabase
+    .from('discussion_messages').select('id, thread_id, sender_type, content, content_type, created_at, attachments')
+    .in('thread_id', threads.map(t => t.id)).order('created_at', { ascending: true });
+
+  if (error) return { error: error.message };
+  return { comments: (messages || []).map(m => ({ id: m.id, sender_type: m.sender_type, content: m.content, created_at: m.created_at, attachments: m.attachments })), total: (messages || []).length };
+}
+
 async function getProjectInfo() {
   // Get project details
   const { data: project, error: projectError } = await supabase
@@ -1810,6 +2623,7 @@ async function listTestCases(args: {
   missing_target_route?: boolean;
   limit?: number;
   offset?: number;
+  compact?: boolean;
 }) {
   let query = supabase
     .from('test_cases')
@@ -1856,6 +2670,20 @@ async function listTestCases(args: {
     );
   }
 
+  // Compact: return minimal fields only
+  if (args.compact === true) {
+    return {
+      count: testCases.length,
+      testCases: testCases.map((tc: any) => ({
+        id: tc.id,
+        testKey: tc.test_key,
+        title: tc.title,
+        priority: tc.priority,
+      })),
+      pagination: { limit, offset, hasMore: testCases.length === limit },
+    };
+  }
+
   return {
     count: testCases.length,
     testCases: testCases.map((tc: any) => ({
@@ -2467,8 +3295,8 @@ async function getCoverageMatrix(args: {
   include_execution_data?: boolean;
   include_bug_counts?: boolean;
 }) {
-  const includeExecution = args.include_execution_data !== false;
-  const includeBugs = args.include_bug_counts !== false;
+  const includeExecution = args.include_execution_data === true;
+  const includeBugs = args.include_bug_counts === true;
 
   // Get tracks
   const { data: tracks } = await supabase
@@ -4167,7 +4995,7 @@ async function markFixedWithCommit(args: {
     status: 'resolved',
     resolved_at: new Date().toISOString(),
     resolution_notes: args.resolution_notes || `Fixed in commit ${args.commit_sha.slice(0, 7)}`,
-    notify_tester: args.notify_tester === true, // Opt-in: only notify if explicitly requested
+    notify_tester: args.notify_tester !== false, // Default: notify tester. Pass false to silently resolve.
     code_context: {
       ...existingContext,
       fix: {
@@ -4190,7 +5018,8 @@ async function markFixedWithCommit(args: {
     return { error: error.message };
   }
 
-  const notificationStatus = args.notify_tester
+  const notifyTester = args.notify_tester !== false;
+  const notificationStatus = notifyTester
     ? 'The original tester will be notified and assigned a verification task.'
     : 'No notification sent (silent resolve). A verification task was created.';
 
@@ -4199,7 +5028,7 @@ async function markFixedWithCommit(args: {
     message: `Bug marked as fixed in commit ${args.commit_sha.slice(0, 7)}. ${notificationStatus}`,
     report_id: args.report_id,
     commit: args.commit_sha,
-    tester_notified: args.notify_tester === true,
+    tester_notified: notifyTester,
     next_steps: [
       'Consider running create_regression_test to prevent this bug from recurring',
       'Push your changes to trigger CI/CD',
@@ -5039,16 +5868,20 @@ Which files or areas would you like me to analyze?`;
 async function listTesters(args: {
   status?: string;
   platform?: string;
+  role?: string;
 }) {
   let query = supabase
     .from('testers')
-    .select('id, name, email, status, platforms, tier, assigned_count, completed_count, notes, created_at')
+    .select('id, name, email, status, platforms, tier, assigned_count, completed_count, notes, role, created_at')
     .eq('project_id', currentProjectId)
     .order('name', { ascending: true });
 
   if (args.status) {
     query = query.eq('status', args.status);
   }
+  if (args.role) {
+    query = query.eq('role', args.role);
+  }
 
   const { data, error } = await query;
 
@@ -5077,6 +5910,7 @@ async function listTesters(args: {
       assignedCount: t.assigned_count,
       completedCount: t.completed_count,
       notes: t.notes,
+      role: t.role,
     })),
   };
 }
@@ -5321,6 +6155,22 @@ async function assignTests(args: {
     status: 'pending',
   }));
 
+  // Helper: after assignments change, sync the test run's total_tests counter
+  async function syncRunCounter() {
+    if (!args.test_run_id) return;
+    const { count } = await supabase
+      .from('test_assignments')
+      .select('id', { count: 'exact', head: true })
+      .eq('test_run_id', args.test_run_id)
+      .eq('project_id', currentProjectId);
+    if (count !== null) {
+      await supabase
+        .from('test_runs')
+        .update({ total_tests: count })
+        .eq('id', args.test_run_id);
+    }
+  }
+
   // Insert — use upsert-like approach: insert and handle conflicts
   const { data: inserted, error: insertErr } = await supabase
     .from('test_assignments')
@@ -5349,6 +6199,8 @@ async function assignTests(args: {
         }
       }
 
+      await syncRunCounter();
+
       return {
         success: true,
         created: created.length,
@@ -5361,6 +6213,8 @@ async function assignTests(args: {
     return { error: insertErr.message };
   }
 
+  await syncRunCounter();
+
   return {
     success: true,
     created: (inserted || []).length,
@@ -5371,6 +6225,82 @@ async function assignTests(args: {
   };
 }
 
+async function unassignTests(args: {
+  assignment_ids: string[];
+}) {
+  if (!args.assignment_ids || args.assignment_ids.length === 0) {
+    return { error: 'At least one assignment_id is required' };
+  }
+  if (args.assignment_ids.length > 50) {
+    return { error: 'Maximum 50 assignments per unassign batch' };
+  }
+  const invalidIds = args.assignment_ids.filter(id => !isValidUUID(id));
+  if (invalidIds.length > 0) {
+    return { error: `Invalid UUID(s): ${invalidIds.join(', ')}` };
+  }
+
+  // Verify assignments exist and belong to this project
+  const { data: existing, error: lookupErr } = await supabase
+    .from('test_assignments')
+    .select('id, test_run_id, test_case:test_cases(test_key, title), tester:testers(name)')
+    .eq('project_id', currentProjectId)
+    .in('id', args.assignment_ids);
+
+  if (lookupErr) return { error: lookupErr.message };
+
+  if (!existing || existing.length === 0) {
+    return { error: 'No matching assignments found in this project' };
+  }
+
+  const foundIds = new Set(existing.map((a: any) => a.id));
+  const notFound = args.assignment_ids.filter(id => !foundIds.has(id));
+
+  // Delete the assignments
+  const { error: deleteErr } = await supabase
+    .from('test_assignments')
+    .delete()
+    .eq('project_id', currentProjectId)
+    .in('id', args.assignment_ids);
+
+  if (deleteErr) return { error: deleteErr.message };
+
+  // Sync run counters for any affected test runs
+  const affectedRunIds = [...new Set(existing.filter((a: any) => a.test_run_id).map((a: any) => a.test_run_id))];
+  for (const runId of affectedRunIds) {
+    const { count } = await supabase
+      .from('test_assignments')
+      .select('id', { count: 'exact', head: true })
+      .eq('test_run_id', runId)
+      .eq('project_id', currentProjectId);
+    if (count !== null) {
+      await supabase.from('test_runs').update({ total_tests: count }).eq('id', runId);
+    }
+  }
+
+  const deleted = existing.map((a: Record<string, unknown>) => {
+    const tc = a.test_case as Record<string, string> | null;
+    const tester = a.tester as Record<string, string> | null;
+    return {
+      id: a.id as string,
+      testKey: tc?.test_key || null,
+      testTitle: tc?.title || null,
+      testerName: tester?.name || null,
+    };
+  });
+
+  const firstKey = deleted[0]?.testKey;
+
+  return {
+    success: true,
+    deletedCount: existing.length,
+    deleted,
+    notFound: notFound.length > 0 ? notFound : undefined,
+    message: existing.length === 1
+      ? `Removed 1 assignment${firstKey ? ` (${firstKey})` : ''}`
+      : `Removed ${existing.length} assignment(s)`,
+  };
+}
+
 async function getTesterWorkload(args: {
   tester_id: string;
 }) {
@@ -5459,6 +6389,7 @@ async function createTester(args: {
   platforms?: string[];
   tier?: number;
   notes?: string;
+  role?: string;
 }) {
   if (!args.name || args.name.trim().length === 0) {
     return { error: 'Tester name is required' };
@@ -5489,8 +6420,9 @@ async function createTester(args: {
       tier: args.tier ?? 1,
       notes: args.notes?.trim() || null,
       status: 'active',
+      role: args.role || 'tester',
     })
-    .select('id, name, email, status, platforms, tier, notes, created_at')
+    .select('id, name, email, status, platforms, tier, notes, role, created_at')
     .single();
 
   if (error) {
@@ -5510,6 +6442,7 @@ async function createTester(args: {
       platforms: data.platforms,
       tier: data.tier,
       notes: data.notes,
+      role: data.role,
       createdAt: data.created_at,
     },
     message: `Tester "${data.name}" added to the project. Use assign_tests to give them test cases.`,
@@ -5796,6 +6729,8 @@ async function getTesterLeaderboard(args: {
 
 async function exportTestResults(args: {
   test_run_id: string;
+  compact?: boolean;
+  limit?: number;
 }) {
   if (!isValidUUID(args.test_run_id)) {
     return { error: 'Invalid test_run_id format' };
@@ -5844,27 +6779,42 @@ async function exportTestResults(args: {
   const passCount = all.filter(a => a.status === 'passed').length;
   const failCount = all.filter(a => a.status === 'failed').length;
 
+  const testRunInfo = {
+    id: run.id,
+    name: run.name,
+    description: run.description,
+    status: run.status,
+    startedAt: run.started_at,
+    completedAt: run.completed_at,
+    createdAt: run.created_at,
+  };
+
+  const summaryInfo = {
+    totalAssignments: all.length,
+    passed: passCount,
+    failed: failCount,
+    blocked: all.filter(a => a.status === 'blocked').length,
+    skipped: all.filter(a => a.status === 'skipped').length,
+    pending: all.filter(a => a.status === 'pending').length,
+    inProgress: all.filter(a => a.status === 'in_progress').length,
+    passRate: all.length > 0 ? Math.round((passCount / all.length) * 100) : 0,
+  };
+
+  // Compact: return test run info + summary only, no assignments array
+  if (args.compact === true) {
+    return { testRun: testRunInfo, summary: summaryInfo };
+  }
+
+  // Apply limit (default: 100, max: 500)
+  const assignmentLimit = Math.min(Math.max(args.limit ?? 100, 1), 500);
+  const limitedAssignments = all.slice(0, assignmentLimit);
+
   return {
-    testRun: {
-      id: run.id,
-      name: run.name,
-      description: run.description,
-      status: run.status,
-      startedAt: run.started_at,
-      completedAt: run.completed_at,
-      createdAt: run.created_at,
-    },
-    summary: {
-      totalAssignments: all.length,
-      passed: passCount,
-      failed: failCount,
-      blocked: all.filter(a => a.status === 'blocked').length,
-      skipped: all.filter(a => a.status === 'skipped').length,
-      pending: all.filter(a => a.status === 'pending').length,
-      inProgress: all.filter(a => a.status === 'in_progress').length,
-      passRate: all.length > 0 ? Math.round((passCount / all.length) * 100) : 0,
-    },
-    assignments: all.map((a: any) => ({
+    testRun: testRunInfo,
+    summary: summaryInfo,
+    assignmentsReturned: limitedAssignments.length,
+    assignmentsTotal: all.length,
+    assignments: limitedAssignments.map((a: any) => ({
       id: a.id,
       status: a.status,
       assignedAt: a.assigned_at,
@@ -5999,6 +6949,12 @@ async function main() {
         case 'get_report_context':
           result = await getReportContext(args as any);
           break;
+        case 'add_report_comment':
+          result = await addReportComment(args as any);
+          break;
+        case 'get_report_comments':
+          result = await getReportComments(args as any);
+          break;
         case 'get_project_info':
           result = await getProjectInfo();
           break;
@@ -6112,6 +7068,9 @@ async function main() {
         case 'assign_tests':
           result = await assignTests(args as any);
           break;
+        case 'unassign_tests':
+          result = await unassignTests(args as any);
+          break;
         case 'get_tester_workload':
           result = await getTesterWorkload(args as any);
           break;
@@ -6147,6 +7106,32 @@ async function main() {
         case 'get_current_project':
           result = getCurrentProject();
           break;
+        // === TEST EXECUTION INTELLIGENCE ===
+        case 'get_test_impact':
+          result = await getTestImpact(args as any);
+          break;
+        case 'get_flaky_tests':
+          result = await getFlakyTests(args as any);
+          break;
+        case 'assess_test_quality':
+          result = await assessTestQuality(args as any);
+          break;
+        case 'get_test_execution_summary':
+          result = await getTestExecutionSummary(args as any);
+          break;
+        case 'check_test_freshness':
+          result = await checkTestFreshness(args as any);
+          break;
+        case 'get_untested_changes':
+          result = await getUntestedChanges(args as any);
+          break;
+        // === AUTO-MONITORING TOOLS ===
+        case 'get_auto_detected_issues':
+          result = await getAutoDetectedIssues(args as any);
+          break;
+        case 'generate_tests_from_errors':
+          result = await generateTestsFromErrors(args as any);
+          break;
         default:
           return {
             content: [{ type: 'text', text: `Unknown tool: ${name}` }],