Spaces:
Sleeping
Sleeping
Commit History
-
docs: Fix field names in mcp_data_structure.md to match implementation
- macro-basket: cpi_yoy → cpi_inflation, unemployment_rate → unemployment, fed_funds_rate → interest_rate - valuation-basket: price_to_book → pb_ratio, price_to_sales → ps_ratio - Added notes explaining field meanings and fallback behavior Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>4edd0ae
-
Refactor: Source-keyed output format across all MCP servers
- MCPs now emit {source: {data: ...}} format directly - Simplify mcp_client.py normalizers to pass-through - Update news-basket to source-keyed structure (tavily, nyt, newsapi) - Update data schema docs to reflect new output format - Streamline fundamentals orchestrator Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>50c0d03 -
Cleanup: Remove obsolete docs, add new schemas, deduplicate configs
- Remove 24 obsolete documentation files (old schemas, test reports) - Add CLAUDE.md project instructions - Add new docs: mcp_output_schemas.md, nvidia_mcp_output_sample.md, sec_edgar_data_schema.md - Deduplicate company_name_filters.py (use root /configs/ version) - Update news-basket to import from shared configs Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>332e21c
-
Feat: Industry-specific metrics via SIC code detection + company address
- Add SIC-to-sector mapping for 13 industries (config.py) - Add industry-specific XBRL concept mappings for: Insurance, Banks, REITs, Oil & Gas, Utilities, Technology, Healthcare, Retail, Financials, Industrials, Transportation, Materials, Mining - Add 60+ industry-specific fields to ParsedFinancials schema - Extract industry metrics based on detected sector (parser.py) - Add business_address extraction from SEC EDGAR submissions - Include company info (with address) in get_all_sources_fundamentals - Add Company Info section to E2E test report Tested with CVX (Oil & Gas), V (Financials), L (Insurance) Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>ab2f4fb
-
Add human-readable output format documentation
Added example output tables for quantitative and qualitative data using markdown format matching mcp_test_report format. Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>f28dae4 -
Add output format documentation to metrics schema
Documents the JSON output structure for: - Aggregated response from fetch_all_research_data - Individual MCP outputs (fundamentals, valuation, volatility, macro, news, sentiment) - Streamed partial_metrics format Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>7b2e271 -
Add schema normalizers for MCP output consistency
- Add _normalize_fundamentals(), _normalize_valuation(), _normalize_volatility(), _normalize_macro() - Flatten sources wrapper to top-level keys (sec_edgar, yahoo_finance, etc.) - Rename financials to fundamentals for consistency across codebase - Add docs: metrics_schema_emitted.md, schema_normalization.md Co-Authored-By: Claude Opus 4.5 <noreply@anthropic.com>26c5c2f