v1.7.3 — AI Diagnosis Overhaul + Auto-Collect Agents

April 6, 2026 · 6 min read

The AI Diagnosis feature has been completely revamped with 3 export formats (DOCX/Markdown/PDF), scheduled reports (weekly/biweekly/monthly), and 6 Auto-Collect agents that automatically gather data from multiple sources. The EKS pages also received major UX improvements including cluster access management, Describe panel, and Service Resources tab.

Key Changes

AI Diagnosis Renewal · DOCX/MD/PDF Export · Scheduled Reports · 6 Auto-Collect Agents · FinOps MCP 5 Tools · EKS Cluster Access Management · Powerpipe Exit Code 2 Handling

AI Diagnosis Renewal

The route has been renamed from /diagnosis to /ai-diagnosis, and the entire UI has been revamped.

Before / After

Item	Before (v1.7.2)	After (v1.7.3)
Route	`/diagnosis`	`/ai-diagnosis`
Name	Comprehensive Diagnosis	AI Diagnosis
Export	PPTX	DOCX + Markdown + PDF
Scheduled Reports	—	Weekly/Biweekly/Monthly
S3 Upload	—	Auto-upload DOCX + MD

3 Export Formats

DOCX — Professional A4 report via docx package:

Cover page (title, health score, account alias, date)
Table of contents (hyperlinks, Heading 1-3)
Section-specific accent colors (cost=green, security=red, network/EKS=purple, idle/MSK=orange)
Markdown parsing (headings, tables, lists, blockquotes, inline code/bold/italic)

Markdown — All sections concatenated into a single .md file.

PDF — Dedicated print page at /ai-diagnosis/report:

White background A4 layout with CSS @media print page breaks
"Print / Save as PDF" button triggers browser Print-to-PDF

// src/lib/report-docx.ts (key structure)
const doc = new Document({
  sections: [
    coverPage(title, healthScore, accountAlias, date),
    tableOfContents(),
    ...sections.map(s => sectionPage(s, accentColor)),
  ],
});
const buffer = await Packer.toBuffer(doc);

Scheduled Reports

Item	Description
Frequency	Weekly / Biweekly / Monthly
Timezone	KST (UTC+9)
Configuration	Day of week (0-6), day of month (1-28), hour (0-23)
Storage	`data/report-schedule.json`
Check interval	Every 5 minutes

// src/lib/report-scheduler.ts
type Schedule = {
  enabled: boolean;
  frequency: 'weekly' | 'biweekly' | 'monthly';
  dayOfWeek?: number;   // 0=Sun ~ 6=Sat
  dayOfMonth?: number;  // 1~28
  hour: number;         // KST
  language: 'ko' | 'en';
};

Toggle ON/OFF in the UI, set frequency/day/hour/language, and reports are automatically generated and uploaded to S3.

S3 Upload

DOCX and Markdown files are automatically uploaded to awsops-deploy-{accountId}/reports/. Download via Presigned URLs valid for 7 days. S3 PutObject/GetObject permissions are automatically added to the CDK stack.

6 Auto-Collect Agents

Dedicated agents that automate data collection for the AI Diagnosis report. Each agent collects data from multiple sources in parallel and streams live status updates.

Agent	Sources	Collected Data
EKS Optimize	Prometheus + K8s + Steampipe	CPU/memory usage, throttling, pod restarts, HTTP 5xx, node utilization
DB Optimize	Steampipe + CloudWatch	RDS/ElastiCache/OpenSearch discovery + metric-based rightsizing
MSK Optimize	Steampipe + CloudWatch + Prometheus	Broker metrics, message throughput, consumer lag, partition count
Idle Scan	Steampipe	Unattached EBS, gp2 volumes, unused EIPs, stopped EC2, 90+ day snapshots, unused SGs + cost estimation
Trace Analyze	Tempo/Jaeger + Prometheus	Per-service error traces, slow traces (>500ms)
Incident	CloudWatch + K8s + Prometheus	ALARM state alarms, Warning events, HTTP 5xx spikes, CPU spikes, memory pressure

// src/lib/collectors/types.ts
export interface Collector {
  collect(send: SendFn, accountId?: string, isEn?: boolean): Promise<CollectorResult>;
  formatContext(data: CollectorResult): string;
  analysisPrompt: string;
  displayName: string;
}

Collector Pattern

All agents implement the same Collector interface. collect() uses Promise.allSettled for parallel multi-source collection and delivers progress via SendFn callback. To add a new agent, just implement this interface.

FinOps MCP Lambda

5 FinOps tools added to the Cost Gateway.

Tool	Source	Description
`get_rightsizing_recommendations`	Compute Optimizer	EC2/RDS/ECS/Lambda rightsizing + monthly savings
`get_savings_plans_recommendations`	Cost Explorer	Savings Plans purchase recommendations (1yr/3yr, No Upfront)
`get_reserved_instance_recommendations`	Cost Explorer	RI purchase recommendations (EC2/RDS/ElastiCache/Redshift)
`get_cost_optimization_hub_recommendations`	Cost Optimization Hub	Unified optimization recommendations (Rightsize/Stop/Upgrade/SP)
`get_trusted_advisor_cost_checks`	Trusted Advisor	Cost optimization category checks + flagged resources

All tools support cross-account queries via the target_account_id parameter.

EKS/K8s Major UX Improvements

Cluster Access Management

Feature	Description
Access Status Badge	Per-cluster "Access Entry" registration status
One-Click Registration	"Register ViewPolicy" button auto-creates Access Entry + Policy
Connection Detection	Detects actual connectivity by checking `context_name` in K8s query results
CLI Guide	Shows EC2 Role ARN-based registration commands for unconnected clusters

New Features

Feature	File	Description
Describe Panel	`K9sDetailPanel`	JSONB detail view (labels, annotations, containers, volumes, conditions)
Service Resources Tab	`k8s.ts`	Per-service CPU/Memory bar charts (Pod selector join)
Cluster Filter	`k8s/page.tsx`	Click cluster cards to filter nodes/pods/events/services
StatsCard Navigation	`StatsCard.tsx`	Nodes→`/k8s/nodes`, Pods→`/k8s/pods`, etc. click navigation
SQL Consolidation	`queries/k8s.ts`	Inline SQL moved to centralized query file + cache warmer

Bug Fixes

external_ip → external_ips column name fix (services 0/0 bug)
EC2 IAM Role ARN truncation fix (SQL || operator filter conflict)
Prevent cluster filter toggle on Register click (stopPropagation)
Post-registration polling for connection verification (5 attempts × 3s)

Benchmark: Powerpipe Exit Code 2 Handling

Powerpipe exits with exit code 2 when benchmark controls have alarms. This is expected behavior, but the previous code treated it as an error.

Item	Before	After
Criteria	exit code === 0	Output file exists + size > 0
Exit code 2	Error	Normal

# Ignore exit code, validate by output file
powerpipe benchmark run ... > "${tmpFile}" 2>"${errorFile}"; \
if [ -s "${tmpFile}" ]; then
  mv "${tmpFile}" "${resultFile}" && echo "done" > "${statusFile}";
else
  echo "error" > "${statusFile}";
fi

Infrastructure: S3 Report Upload + update-infra Script

S3 Policy

S3 permissions added to the EC2 role in the CDK stack:

// infra-cdk/lib/awsops-stack.ts
ec2Role.addToPolicy(new iam.PolicyStatement({
  actions: ['s3:PutObject', 's3:GetObject'],
  resources: [`arn:aws:s3:::awsops-deploy-${this.account}/reports/*`],
}));

update-infra Script

00-update-infra.sh script for safe CDK stack updates without EC2 replacement:

Read existing stack parameters — Preserve current CloudFormation values
Reconstruct CDK context — Re-derive from live infrastructure (VPC, TGW, domain)
cdk diff preview — Warn if EC2 would be replaced
Deploy — Safe deployment with --no-rollback

Version Comparison

Item	v1.7.2	v1.7.3	Change
Diagnosis route	`/diagnosis`	`/ai-diagnosis`	Renamed
Export formats	PPTX	DOCX + MD + PDF	3 formats
Scheduled reports	—	Weekly/Biweekly/Monthly	New
Auto-Collect	—	6 agents	New
FinOps tools	—	5 MCP tools	New
EKS access	Manual	One-click registration	Improved
Benchmark	Exit code dependent	Output file validation	Fixed

Key File Changes

File	Change
`src/app/ai-diagnosis/page.tsx`	Diagnosis page renewal (route change, schedule UI)
`src/app/ai-diagnosis/report/page.tsx`	PDF print-friendly page
`src/app/api/report/route.ts`	Report generation API (15 sections, S3 upload)
`src/lib/report-docx.ts`	DOCX generator (cover, TOC, section styles)
`src/lib/report-scheduler.ts`	Scheduled report scheduler (5-min check)
`src/lib/report-prompts.ts`	15 section-specific Bedrock analysis prompts
`src/lib/report-generator.ts`	Data collection orchestrator
`src/lib/collectors/*.ts`	6 Auto-Collect agents
`agent/lambda/aws_finops_mcp.py`	FinOps MCP Lambda (5 tools)
`src/app/api/benchmark/route.ts`	Powerpipe exit code 2 handling
`src/lib/queries/k8s.ts`	K8s SQL query consolidation + Service Resources
`src/components/dashboard/StatsCard.tsx`	Click navigation (`href` prop)
`infra-cdk/lib/awsops-stack.ts`	S3 reports/* policy
`scripts/00-update-infra.sh`	Safe infrastructure update script

AI Diagnosis Renewal​

Before / After​

3 Export Formats​

Scheduled Reports​

6 Auto-Collect Agents​

FinOps MCP Lambda​

EKS/K8s Major UX Improvements​

Cluster Access Management​

New Features​

Bug Fixes​

Benchmark: Powerpipe Exit Code 2 Handling​

Infrastructure: S3 Report Upload + update-infra Script​

S3 Policy​

update-infra Script​

Version Comparison​

Key File Changes​