aaaaqwq/weekly-review-run

Weekly Review - Run Agent

Generate weekly summary report manually or test the automated agent.

When to use

• "generate weekly review"
• "run weekly report"
• "show report for the week"
• "test weekly review agent"

What it does

Generates a comprehensive weekly summary from:

• Metrics: hours, tokens, tasks completed
• Breakdown by projects and activity types
• CRM activities
• Learnings for the week
• Blocked tasks
• Upcoming deadlines
• AI insights (achievements, blockers, recommendations)

How to execute

Option 1: Generate report (recommended)

cd $AGENTS_PATH/weekly-review
python3 weekly_review_agent.py --dry-run

Result:

• Generates report in $PROJECT_ROOT/reports/weekly/
• Shows preview notification
• Does NOT send to Telegram

Option 2: Full run

cd $AGENTS_PATH/weekly-review
python3 weekly_review_agent.py

Result:

• Generates report
• Sends notification to Telegram

Option 3: Without AI (fast)

cd $AGENTS_PATH/weekly-review
python3 weekly_review_agent.py --dry-run --no-ai

Result:

• Data only, no AI insights
• Faster execution

Option 4: For a specific date

cd $AGENTS_PATH/weekly-review
python3 weekly_review_agent.py --date 2026-02-14 --dry-run

Result:

• Report for the week ending 2026-02-14

Option 5: Telegram test

cd $AGENTS_PATH/weekly-review
python3 weekly_review_agent.py --notify-test

Output

Report file

Location: $PROJECT_ROOT/reports/weekly/weekly_YYYY-MM-DD.md

Sections:

1. Executive Summary (hours, tokens, tasks)
2. Hours by Project
3. Hours by Activity Type
4. Completed Tasks
5. CRM Activities
6. Learnings
7. Blocked Tasks
8. Upcoming Deadlines
9. AI Insights (achievements, blockers, recommendations)
10. Next Week Focus

Telegram Notification

Target: Info channel (currently "me" - Saved Messages)

Format:

📊 Weekly Review Ready

📅 Period: 2026-02-03 to 2026-02-09
⏱ Hours: 12.5h
✅ Tasks: 8 completed
🎯 Tokens: 45,000

🏆 Top project: lung-xray-analyzer (6.2h)
⚠️ Blocked tasks: 2
📆 Upcoming deadlines: 3

📄 Full report: weekly_2026-02-09.md

Troubleshooting

Error: files missing

Check if data exists:

ls -lh $PM_PATH/pm_*.csv
ls -lh $CRM_PATH/activities.csv

Claude CLI error

Test:

echo "test" | claude -p --model haiku --output-format text

If not working -- agent will use fallback (without AI insights).

Telegram not working

Test:

python3 weekly_review_agent.py --notify-test

Check tg-tools auth:

ls $SALES_PATH/telegram/sessions/

Empty report

If there is no data for the week -- agent will generate a minimal report with the message "No activity this week".

Automation

Agent runs automatically every Friday at 18:00 via launchd.

Status:

launchctl list | grep weekly-review

Logs:

tail -f $AGENTS_PATH/logs/weekly_review_stderr.log

Recent runs:

cat $AGENTS_PATH/logs/weekly_review.log | jq '.'

Testing

Run test suite:

cd $AGENTS_PATH/weekly-review
python3 test_weekly_review.py

Related skills

• daily-briefing — daily tactical brief (daily vs weekly strategic)
• task-prioritization — uses weekly insights for prioritization
• log-activity — logs CRM activities that appear in weekly review

Notes

• Read-only: Agent only reads data, does not modify anything
• Idempotent: Safe to run multiple times
• No approvals: Fully automatic, informational only
• Graceful degradation: If a section fails -- continues with available data