macro-briefing

description	SKILL_DIR="$(dirname "$(realpath "$0")")/.." # resolve from scripts/
name	macro-briefing
category	trading-finance
maturity	stable
tags	["yfinance","economic-calendar","yield-curve","vix","sqlite"]

Macro Briefing

Quick Start

Collect + brief (one-liner)

SKILL_DIR="$(dirname "$(realpath "$0")")/.."  # resolve from scripts/
python3 scripts/collect.py && python3 scripts/briefing.py --format telegram

Collect only (stores to DB)

python3 scripts/collect.py --db data/macro.db

Generate briefing from latest DB snapshot

python3 scripts/briefing.py --format telegram --session europe

Scripts

scripts/collect.py — Data Collector

Fetches all market data via yfinance + economic calendar via FairEconomy API. Stores in SQLite (data/macro.db by default). No API keys required.

Flags:

• --db PATH — custom DB path (default: data/macro.db)
• --no-db — skip DB storage, just collect
• --json — output raw JSON to stdout

Dependency: pip3 install yfinance (pandas, curl_cffi auto-installed)

scripts/briefing.py — Briefing Formatter

Reads latest snapshot from DB (or JSON stdin) and formats for delivery.

Flags:

• --format telegram|plain|json — output format
• --session asia|europe|us — filter calendar events by session window
• --db PATH — custom DB path
• --stdin — read JSON from stdin instead of DB
• --json-input PATH — read from JSON file

Formats:

• telegram — Markdown with emoji, ready for Telegram message
• plain — stripped plain text
• json — structured JSON for programmatic use by other skills

scripts/query.py — Database Query API

Historical queries for analysis, backtesting, and other skills.

Commands:

• ticker DXY --days 30 [--csv] — single ticker history
• yields --days 60 [--csv] — yield curve history
• spread 10y3m --days 180 — spread history
• snapshot [--date 2026-03-01] — full snapshot (latest or by date)
• calendar [--country USD] [--impact High] [--days 7] — economic events
• summary --days 5 — summary stats (mean/min/max/trend)
• export --days 365 --csv — full export for backtesting

Database

SQLite at data/macro.db. Schema in scripts/collect.py init_db().

Tables:

• snapshots — full JSON per run (for complete reconstruction)
• market_series — flat rows: (collected_at, ticker, close, change_pct, ...) — use this for queries
• yield_curve — dedicated curve + spread columns
• calendar_events — economic events with country/impact

Indexes: ticker+date on market_series, date on snapshots, date on calendar.

Automated Delivery

Set up via OpenClaw cron for pre-session Telegram delivery. See references/trading-sessions.md for schedule.

Example cron task (Europe open, 06:45 UTC Mon-Fri):

Task: Run macro-briefing collect.py then briefing.py --format telegram --session europe. Send the output to Telegram.

Integration with Other Skills

Other skills consume macro data in three ways:

1. Run scripts/query.py — CLI for any historical query
2. Read data/macro.db directly via SQLite — full programmatic access
3. Run scripts/briefing.py --format json — structured current snapshot

Example from another skill:

# Get current VIX regime
python3 /path/to/macro-briefing/scripts/query.py snapshot | jq '.derived.vix_regime'

# Export 90 days of DXY for backtest
python3 /path/to/macro-briefing/scripts/query.py ticker DXY --days 90 --csv > dxy_90d.csv

Adding Tickers

Edit TICKERS dict in scripts/collect.py. Any valid Yahoo Finance symbol works. Run collect.py once to populate the new ticker in DB.

References

• references/data-sources.md — all tickers, data sources, derived indicators, DB schema
• references/trading-sessions.md — session windows, delivery schedule, cron examples, key release times