data-us
Installation
SKILL.md
data-us
Layer 1 data skill for the US market in the v2.0.0 three-layer architecture
(Data / Analysis / Report). This skill is fetch-only — it does not
analyze, score, or compose narrative. All output is structured JSON intended
for direct handoff to a Layer 2 analysis-* skill or a Layer 3 report-*
orchestrator.
Sources
| Layer | Source | Script | Returns |
|---|---|---|---|
| Snapshot / multiples | yfinance (unofficial) | ${CLAUDE_SKILL_DIR}/scripts/yfinance_client.py |
OHLCV history, info dict (PE/PB/marketCap/beta/dividendYield/sector etc.) |
| Fundamental | SEC EDGAR (data.sec.gov, official Tier A) |
${CLAUDE_SKILL_DIR}/scripts/sec_edgar_client.py |
10-K / 10-Q / 8-K filings index, XBRL us-gaap facts, Item-section narrative |
| Macro | FRED CSV | ${CLAUDE_SKILL_DIR}/scripts/fred_client.py |
Time-series for rates / inflation / growth / nowcast / real-rates / swap-spreads |
All three client scripts are byte-identical copies of the canonical
versions under investing-toolkit/scripts/ (drift is enforced by CI; see
repo scripts/sync-check.sh).