Weekly Summary

Weekly Summary - AI-Powered Multi-Repo Analysis

Generate comprehensive weekly engineering summaries with:

• All branch analysis (including feature branches)
• AI-powered summarization per repository
• Alpha deployment detection (feature → dev merges)
• Business-focused aggregation

Execution Flow

When /weekly_summary is invoked, follow these steps:

Step 1: Calculate This Week's Date Range

python3 -c "
from datetime import datetime, timedelta
today = datetime.now()
monday = today - timedelta(days=today.weekday())
sunday = monday + timedelta(days=6)
print(f'{monday.strftime(\"%Y-%m-%d\")}|{sunday.strftime(\"%Y-%m-%d\")}')"

Store the output as MONDAY and SUNDAY.

Step 2: Collect Commits from ALL Branches

bash .claude/skills/weekly_summary/collect_all_branches.sh "$MONDAY" "$SUNDAY"

This creates /tmp/weekly_commits_by_repo/.txt files with:

• All commits from all branches (not just main/dev)
• Active branch list
• Complete commit messages with bodies

Step 3: Spawn AI Agents for Each Repository

For each repository with activity, spawn the weekly-repo-analyzer custom agent:

Task(
  subagent_type="weekly-repo-analyzer",
  model="haiku",
  description="Analyze {repo_name} weekly commits",
  prompt="Analyze the commits in /tmp/weekly_commits_by_repo/{repo_file}.txt for {repo_name} covering {MONDAY} to {SUNDAY}."
)

Run all 12 agents in parallel (use multiple Task tool calls in one response for speed).

The custom agent has built-in knowledge of:

• DataStreaming platform architecture
• Deployment stages (dev=alpha, staging, main=prod)
• Initiative keywords and business value extraction
• Component health assessment logic

Simple Agent Prompt

Since the weekly-repo-analyzer agent has all the context built-in, just tell it the file path:

Analyze the commits in /tmp/weekly_commits_by_repo/{repo_file}.txt for {repo_name}.

Week: {MONDAY} to {SUNDAY}

Output valid JSON with initiatives, deployments, bugs fixed, and feature branch activity.

The agent will automatically:

• Aggregate commits into business initiatives
• Detect alpha/staging/production deployments
• Extract business value and impact
• Assess component health
• Identify feature branch work in progress
• Output structured JSON

Step 4: Aggregate All Agent Results

After all agents complete:

1. Read all agent JSON outputs from Task results
2. Save each to /tmp/weekly_agent_outputs/{repo}_analysis.json using Bash:

# Use cat with heredoc or echo to write JSON files
cat > /tmp/weekly_agent_outputs/web-app_analysis.json  /tmp/weekly_agent_outputs/repo_analysis.json

IMPORTANT: Use Bash (cat/echo) instead of Write tool to avoid read-before-write requirements for new files.

1. Run aggregation:

python3 .claude/skills/weekly_summary/orchestrate_summary.py "$MONDAY" "$SUNDAY" --aggregate-only

This generates the final business summary with:

• Executive overview
• Component status table
• Alpha Deployments section (features in testing)
• Top initiatives (AI-aggregated)
• Platform health

Step 5: Save Reports

ai_docs/weekly-summaries/
├── YYYY-MM-DD-business-summary.md  ← Executive report
└── YYYY-MM-DD-technical-summary.md ← Engineering details

Example Agent Spawning

Spawn all 12 repo agents in parallel:

// In one response, call Task tool 12 times:

Task(subagent_type="Explore", description="Analyze web-app", prompt="...")
Task(subagent_type="Explore", description="Analyze workers", prompt="...")
Task(subagent_type="Explore", description="Analyze data-cluster", prompt="...")
// ... for all 12 repos

Repositories to Analyze

• web-app (Backend + Frontend)
• workers (Job processing)
• data-pipelines (Argo workflows)
• data-cluster (FastAPI data API)
• data-cluster-operator (K8s operator)
• data-cluster-helm (Infrastructure Helm)
• k8s-charts (Platform deployment)
• skupper-gateway (Networking)
• MCPs/mcp-base
• MCPs/mcp-boilerplate
• MCPs/mcp-datacluster
• MCPs/mcp-openaire

Key Innovations

✅ Captures ALL branch activity (not just merged) ✅ AI understands context (better than keyword matching) ✅ Aggregates into initiatives (not commit lists) ✅ Detects alpha deployments (feature → dev merges) ✅ Parallel processing (12 agents at once) ✅ Business value focused (not just technical changes)

Expected Output

Business Summary

# Weekly Engineering Update
## 2026-02-02 to 2026-02-08

## Executive Overview
**Focus**: Platform Observability (51 changes, 11 components)
**Health**: Strong
**Shipped**: 90 features, 115 fixes

## Component Status
| Component | Status | Summary |
|---|---|---|
| Backend API | healthy | 20 features, 30 fixes |
| Infrastructure | healthy | 15 features, 12 fixes |

## Alpha Deployments (Features in Testing)

### Uptime Kuma Status Page
**Repo**: k8s-charts • **Date**: 2026-02-04
**Status**: Deployed to alpha (dev branch)
**Impact**: Live status page at status.alien.club

### Public Dataplane Sharing
**Repo**: web-app • **Date**: 2026-02-03
**Status**: Deployed to alpha (dev branch)
**Impact**: Cross-organization data sharing capability

## Top Initiatives
1. Platform Observability (51 changes)
2. Vector Search & AI (26 changes)
3. Infrastructure Monitoring (12 changes) ← NEW!

Troubleshooting

If agents fail: Check /tmp/weekly_commits_by_repo/ for commit files If no alpha deployments detected: Check for "Merge" commits to "dev" branch If analysis incomplete: Ensure Task tool has access to file reads

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: the-alien-club
• Source: the-alien-club/claude-workflows
• License: MIT

Install and usage instructions live in the source repository linked above.