Skip to content

REWRITE_PROGRESS.md

Latest commit

 

History

History
699 lines (571 loc) · 45.6 KB

REWRITE_PROGRESS.md

File metadata and controls

699 lines (571 loc) · 45.6 KB
Raw

fmpsdk Rewrite — Implementation Progress

Checklist of all 238 canonical methods from REWRITE_ARCHITECTURE.md §6. Generated from that file — if the catalog changes, regenerate rather than hand-edit the list.

How to use this file

Update the checkbox and status tag as each method is actually built and tested — this file, not memory or conversation history, is the source of truth for "what's left." Commit it alongside the method(s) it tracks, as its own line in the diff.

Status tags (append after the method name once it's past [ ]):

  • [x] unit — implemented, mocked unit test passing, not yet live-verified
  • [x] live — implemented, unit-tested, and live-verified against the real API (Bucket 1)
  • [x] ultimate-pending — implemented and unit-tested; confirmed to 402 on every FMP tier tested so far (see the parenthetical after each line for exactly which tiers)
  • [x] done — implemented, unit-tested, live-verified on some real tier, nothing left. Check the parenthetical: plain done with no tier note means free-tier; (works on Starter tier; ...) etc. means it needed that paid tier or higher
  • [x] untested (community add-on) — implemented and unit-tested (mocked only); gated behind something outside the Free/Starter/Premium/Ultimate ladder entirely (a separate FMP add-on purchase) that this maintainer has decided not to buy. Permanent, not a todo — only a community PR with real verification can move one of these to done.
  • [ ] blocked: <reason> — attempted, hit something unexpected (e.g. a doc/reality mismatch worth flagging), not resolved yet

FMP's real pricing ladder is Free → Starter → Premium → Ultimate (not just Free/Ultimate as earlier notes in this project assumed) — see the dated entry below and [[fmpsdk-rewrite-status]] for how that was discovered. The plan: whenever Dax upgrades to the next paid tier, re-run pytest tests/ultimate/ -m ultimate, move whatever now passes into tests/live/, and repeat at the next tier up. A method still ultimate-pending after a tier's pass hasn't necessarily earned the name "needs Ultimate" — it just hasn't been confirmed working yet at any tier tried so far; the parenthetical always says which tiers it's actually been tested against and failed on.

Bucket note: group-level Bucket 2 flags below are a best-effort carry-over from the original pricing-tier audit earlier in this project, not verified per-method. The actual live-testing discipline (attempt each method as it's built, one fixed cheap test case) is the real source of truth — if a "Bucket 1" method 402s, mark it ultimate-pending and move on; if a "Bucket 2" method turns out to work on the current key, even better.

Progress: 231 / 238 methods done, 7 untested (community add-on — see the 2026-08-24 entry below), 0 left ultimate-pending. This is the full 238/238 catalog now implemented in code, and every single method has now been attempted at least once (unit-tested, and live-tested except where already Bucket-2-confirmed) — every canonical method in REWRITE_ARCHITECTURE.md §6 has a Python implementation and a test result. 2026-08-23's second work session unit-tested and live-tested all 45 indexes-through-technical_indicators methods from the first code-only pass; a third session live-tested market_performance (11/11, no 402s), wrote mocked unit tests for all 79 methods in news/quote/ sec_filings/earnings_transcript/congress/bulk/tipranks (268/268 unit tests passing), then live-tested all 79 of those too — 19 more passed free-tier (fmp_articles; quote/quote_short/aftermarket_quote/aftermarket_trade/ stock_price_change; 10 of 12 sec_filings methods; house_latest/ senate_latest), the remaining 60 402'd and moved to ultimate-pending (earnings_transcript, bulk, and tipranks confirmed fully gated as their pre-flagged Bucket 2 status predicted; news's batch_* quote family and most of congress turned out gated too, despite no Bucket 2 flag in §3.5 — see the per-group notes below).

2026-08-23, later — Dax upgraded Free → Starter. Learned FMP's pricing ladder actually has 4 tiers (Free/Starter/Premium/Ultimate), not the Free/Ultimate binary this project assumed until now. Re-ran all 129 ultimate-pending methods' existing tests/ultimate/ test bodies (no new tests needed to write) against the Starter-tier key: 51 now pass, 78 still 402. All 51 moved into tests/live/ (merged into each group's existing live file, or a new one where the whole group had been gated), tagged done with a "works on Starter tier" note below; the 78 still-gated stay ultimate-pending, now annotated "402 on free tier and on Starter tier" so a future Premium-tier pass knows exactly what's already been ruled out. Full groups that flipped entirely to done: directory (10/10), insider_trades (4/4, closing that group out completely), technical_indicators (9/9, closing that group out completely), sec_filings (2/2, closing that group out completely — all 12 methods now done), calendar's IPOs (3/3), chart's historical_chart (closing that group out completely), economics's economic_calendar (closing that group out completely), search's remaining 4 (closing that group out completely). Partial flips: news (7 of 9 402'd methods now pass — only the 2 press-releases methods still gated), congress (6 of 10 — the symbol/name-scoped trade lookups now work, the 4 senate-profile-shaped methods don't), funds (3 of 9 — the informational ETF endpoints work, the disclosure/ holdings ones don't), company (1 of 3 — mergers_acquisitions_latest only). Groups that stayed fully gated even at Starter: bulk (18), commitment_of_traders (3), earnings_transcript (4), esg (3), institutional_ownership (8), tipranks (7), indexes's constituent-list methods (6), quote's batch_* family (11), statements's TTM/latest (4). Whole test suite re-run after the reorg: 268/268 unit, 160/160 live (includes the 51 newly-moved), 78/78 ultimate still failing as expected — clean split, no regressions. Docstrings for all affected methods/groups updated same session to state the new Starter-tier facts, replacing flat "requires Ultimate" claims (never actually verified against a real Ultimate key) with "requires FMP Premium or Ultimate, not yet confirmed which" wherever a method was still gated.

2026-08-23, same day, later still — Dax upgraded Starter → Premium (750 calls/min at this tier). Re-ran the remaining 78 ultimate-pending methods' existing tests/ultimate/ test bodies against the Premium-tier key: 21 now pass, 57 still 402. Same move/retag/docstring process as the Starter pass. Groups that flipped entirely to done: commitment_of_traders (3/3), company (now 3/3, closing that group out completely — mergers_acquisitions_search and executive_compensation_benchmark were the last 2), congress (now 10/10 + the 2 free-tier -latest listings = 12/12, closing that group out completely), indexes (all 6 constituent-list methods, closing that group out completely), news (now 10/10, closing that group out completely — news_press_releases and news_press_releases_latest were the last 2). Partial: quote (4 of the remaining 7 batch_* methods now pass — batch_quote, batch_quote_short, batch_aftermarket_quote, batch_aftermarket_trade; the other 7, including batch_exchange_quote and all 6 whole-asset-class ones, still gated). Groups that stayed fully gated even at Premium: bulk (18), earnings_transcript (4), esg (3), funds (6 remaining), institutional_ownership (8), statements's TTM/latest (4), tipranks (7) — these 57 now carry "402 on free, Starter, and Premium tiers" and await a future Ultimate-tier pass (the last rung on FMP's ladder, so whatever's still gated there simply is Ultimate-only — no more ambiguity to track past that point). Whole suite re-verified: 268/268 unit, 181/181 live, 57/57 ultimate still failing as expected — clean split, no regressions. Docstrings updated same session for all 6 affected groups/methods.

2026-08-24 — Dax upgraded Premium → Ultimate (3000 calls/min at this tier — FMP's top standard plan). Re-ran the remaining 57 ultimate-pending methods against the Ultimate-tier key: 50 now pass, 7 still 402. Along the way, running bulk's tests for the first time against a real 200 response (rather than an immediate 402) surfaced a real bug that had nothing to do with plan tier: every client.bulk method's real response is CSV (text/csv), not JSON, despite FMP's own docs showing JSON-looking examples — the existing code called Client._get, which parses JSON and raised JSONDecodeError against real bulk data. Fixed by adding Client._get_csv (shares the retry/error-mapping core via _request, same pattern as _get_bytes for financial_reports_xlsx) and switching all 18 bulk methods to it. Field names/order in every existing *BulkResult TypedDict were verified against real CSV headers and all matched exactly — the one exception was profile_bulk, previously typed as the JSON-shaped ProfileResult on the (wrong) belief that it was the one bulk method returning real JSON; it's CSV too, all-str like every other bulk method, so a new ProfileBulkResult TypedDict replaces that reuse. tests/unit/ test_bulk.py's 18 mocks switched from json= to text= CSV fixtures to match. See [[fmpsdk-rewrite-status]] for the full account.

Groups that flipped entirely to done: bulk (all 18, once the CSV fix landed), earnings_transcript (4/4), esg (3/3), institutional_ownership (8/8), funds (now 9/9, closing that group out completely), quote (now 16/16, closing that group out completely — the remaining 7 batch_* methods), statements (now 27/27, closing that group out completely — the TTM/latest_financial_statements quartet). tipranks's 7 methods still 402 even on Ultimate — but this is not a tier gap. FMP's own error message names the real cause: a separate paid add-on ("TipRanks data boost"), purchased independently of the Free/Starter/Premium/Ultimate ladder via the dashboard's Add-ons tab. Retagged accordingly below rather than left implying a 5th tier exists. Whole suite re-verified: 268/268 unit, 231/231 live, 7/7 ultimate (tipranks) still failing as expected — clean split, no regressions. Docstrings updated same session for all 7 affected groups/methods (including a genuine mistake caught and fixed from the Premium-pass session: quote's module docstring had wrongly claimed 7 batch_* methods were "confirmed working on Premium" when they'd actually still been gated at that tier).

This closes out the entire standard-tier testing backlog — 231/238 done, with the remaining 7 blocked on a separate product purchase (TipRanks add-on) rather than a plan tier.

2026-08-24, same day — Dax decided not to buy the TipRanks add-on. The project's testing effort ends here, permanently, not as a "waiting for the next tier" pause: the 7 tipranks methods are retagged untested (community add-on) rather than ultimate-pending, since there's no further FMP purchase planned that would move them. README and fmpsdk/endpoints/tipranks.py's module docstring updated to say so plainly — implemented and unit-tested (mocked) like everything else, but never verified against a real response, and that's expected to stay true unless a user with the add-on sends a PR. There is no more tier-testing work left in this project.

client.search — Identifier lookup (symbol/name/CIK/CUSIP/ISIN) and the screener.

7 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found 4 of 7 methods 402 on the free tier anyway. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done company_screener — company-screener (works on Starter tier; 402 on free tier)
  • done search_cik — search-cik
  • done search_cusip — search-cusip (works on Starter tier; 402 on free tier)
  • done search_exchange_variants — search-exchange-variants (works on Starter tier; 402 on free tier)
  • done search_isin — search-isin (works on Starter tier; 402 on free tier)
  • done search_name — search-name
  • done search_symbol — search-symbol

client.directory — Whole-universe reference lists: symbols, exchanges, sectors, industries, countries.

10 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found all 10 of 10 methods 402 on the free tier — the whole group, not a subset like search's 4/7. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done actively_trading_list — actively-trading-list (works on Starter tier; 402 on free tier)
  • done available_countries — available-countries (works on Starter tier; 402 on free tier)
  • done available_exchanges — available-exchanges (works on Starter tier; 402 on free tier)
  • done available_industries — available-industries (works on Starter tier; 402 on free tier)
  • done available_sectors — available-sectors (works on Starter tier; 402 on free tier)
  • done cik_list — cik-list (works on Starter tier; 402 on free tier)
  • done etf_list — etf-list (works on Starter tier; 402 on free tier)
  • done financial_statement_symbol_list — financial-statement-symbol-list (works on Starter tier; 402 on free tier)
  • done stock_list — stock-list (works on Starter tier; 402 on free tier)
  • done symbol_change — symbol-change (works on Starter tier; 402 on free tier)

client.analyst — Sell-side estimates, ratings, price targets, grades.

8 methods.

  • done analyst_estimates — analyst-estimates
  • done grades — grades
  • done grades_consensus — grades-consensus
  • done grades_historical — grades-historical
  • done price_target_consensus — price-target-consensus
  • done price_target_summary — price-target-summary
  • done ratings_historical — ratings-historical
  • done ratings_snapshot — ratings-snapshot

client.calendar — Date-driven corporate events: dividends, earnings, IPOs, splits.

9 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found the 3 ipos_* methods 402 on the free tier — the whole IPOs category, while dividends/earnings/splits are all free-tier reachable. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done dividends — dividends
  • done dividends_calendar — dividends-calendar
  • done earnings — earnings
  • done earnings_calendar — earnings-calendar
  • done ipos_calendar — ipos-calendar (works on Starter tier; 402 on free tier)
  • done ipos_disclosure — ipos-disclosure (works on Starter tier; 402 on free tier)
  • done ipos_prospectus — ipos-prospectus (works on Starter tier; 402 on free tier)
  • done splits — splits
  • done splits_calendar — splits-calendar

client.chart — Historical price series, EOD and intraday, for every asset class.

5 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found historical_chart (all 6 intraday timeframes) 402s on the free tier while all 4 historical_price_eod_* daily methods work. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done historical_chart — historical-chart/{timeframe} (works on Starter tier; 402 on free tier)
  • done historical_price_eod_dividend_adjusted — historical-price-eod/dividend-adjusted
  • done historical_price_eod_full — historical-price-eod/full
  • done historical_price_eod_light — historical-price-eod/light
  • done historical_price_eod_non_split_adjusted — historical-price-eod/non-split-adjusted

client.company — Company-level reference and profile data, incl. market cap, float, executives, M&A.

17 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found 3 of 17 methods 402 on the free tier. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done company_notes — company-notes
  • done delisted_companies — delisted-companies
  • done employee_count — employee-count
  • done executive_compensation_benchmark — executive-compensation-benchmark (works on Premium tier; 402 on free and Starter tiers)
  • done governance_executive_compensation — governance-executive-compensation
  • done historical_employee_count — historical-employee-count
  • done historical_market_capitalization — historical-market-capitalization
  • done key_executives — key-executives
  • done market_capitalization — market-capitalization
  • done market_capitalization_batch — market-capitalization-batch
  • done mergers_acquisitions_latest — mergers-acquisitions-latest (works on Starter tier; 402 on free tier)
  • done mergers_acquisitions_search — mergers-acquisitions-search (works on Premium tier; 402 on free and Starter tiers)
  • done profile — profile
  • done profile_cik — profile-cik
  • done shares_float — shares-float
  • done shares_float_all — shares-float-all
  • done stock_peers — stock-peers

client.commitment_of_traders — CFTC Commitment-of-Traders reports and analysis.

3 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found all 3 methods 402 on the free tier. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done commitment_of_traders_analysis — commitment-of-traders-analysis (works on Premium tier; 402 on free and Starter tiers)
  • done commitment_of_traders_list — commitment-of-traders-list (works on Premium tier; 402 on free and Starter tiers)
  • done commitment_of_traders_report — commitment-of-traders-report (works on Premium tier; 402 on free and Starter tiers)

client.dcf — Discounted-cash-flow valuations, standard and custom-input.

4 methods.

  • done custom_discounted_cash_flow — custom-discounted-cash-flow
  • done custom_levered_discounted_cash_flow — custom-levered-discounted-cash-flow
  • done discounted_cash_flow — discounted-cash-flow
  • done levered_discounted_cash_flow — levered-discounted-cash-flow

client.economics — Macroeconomic series, treasury rates, economic calendar, risk premium.

4 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found economic_calendar 402s on the free tier while the other 3 work. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done economic_calendar — economic-calendar (works on Starter tier; 402 on free tier)
  • done economic_indicators — economic-indicators
  • done market_risk_premium — market-risk-premium
  • done treasury_rates — treasury-rates

client.esg — ESG disclosures, ratings, and benchmarks.

Bucket 2 (Ultimate-gated), confirmed live: the pricing-audit prediction held — all 3 methods 402 on the free tier. Tests live in tests/ultimate/test_esg.py only.

3 methods.

  • done esg_benchmark — esg-benchmark (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done esg_disclosures — esg-disclosures (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done esg_ratings — esg-ratings (works on Ultimate tier; 402 on free, Starter, and Premium)

client.funds — ETF and mutual-fund composition, info, and N-PORT/13F-style disclosures.

9 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found all 9 methods 402 on the free tier. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

  • done etf_asset_exposure — etf/asset-exposure (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done etf_country_weightings — etf/country-weightings (works on Starter tier; 402 on free tier)
  • done etf_holdings — etf/holdings (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done etf_info — etf/info (works on Starter tier; 402 on free tier)
  • done etf_sector_weightings — etf/sector-weightings (works on Starter tier; 402 on free tier)
  • done funds_disclosure — funds/disclosure (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done funds_disclosure_dates — funds/disclosure-dates (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done funds_disclosure_holders_latest — funds/disclosure-holders-latest (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done funds_disclosure_holders_search — funds/disclosure-holders-search (works on Ultimate tier; 402 on free, Starter, and Premium)

client.statements — Financial statements and everything computed directly from them.

27 methods.

Correction to the group directory's Bucket assignment: not flagged as Bucket 2 in REWRITE_ARCHITECTURE.md §3.5, but live-testing found 4 of 27 methods 402 on the free tier. Reclassified below per the workflow doc's "if a Bucket 1 method 402s, mark it ultimate-pending" rule.

Two doc/reality mismatches found and resolved, not just noted: financial_reports_xlsx really does return binary XLSX bytes (verified via ZIP magic bytes PK\x03\x04) despite a lying application/json content-type header and a docs example that's a copy-paste of the JSON endpoint's — exactly what REWRITE_ARCHITECTURE.md §8.4 predicted and flagged for live verification. financial_reports_json turned out to return a single bare object, not array-wrapped like the other 242 endpoints in the catalog (contradicting FMP's own docs, which show it array-wrapped) — a second, previously-unflagged exception to the List[Dict] response contract. Both are implemented correctly (not forced into the wrong shape) — see client.py's _get_bytes() and financial_reports_json's return type in endpoints/statements.py.

  • done balance_sheet_statement — balance-sheet-statement
  • done balance_sheet_statement_as_reported — balance-sheet-statement-as-reported
  • done balance_sheet_statement_growth — balance-sheet-statement-growth
  • done balance_sheet_statement_ttm — balance-sheet-statement-ttm (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done cash_flow_statement — cash-flow-statement
  • done cash_flow_statement_as_reported — cash-flow-statement-as-reported
  • done cash_flow_statement_growth — cash-flow-statement-growth
  • done cash_flow_statement_ttm — cash-flow-statement-ttm (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done enterprise_values — enterprise-values
  • done financial_growth — financial-growth
  • done financial_reports_dates — financial-reports-dates
  • done financial_reports_json — financial-reports-json (returns a single dict, not array-wrapped — verified live, contradicts FMP's own docs)
  • done financial_reports_xlsx — financial-reports-xlsx (returns raw bytes — verified live per §8.4)
  • done financial_scores — financial-scores
  • done financial_statement_full_as_reported — financial-statement-full-as-reported
  • done income_statement — income-statement
  • done income_statement_as_reported — income-statement-as-reported
  • done income_statement_growth — income-statement-growth
  • done income_statement_ttm — income-statement-ttm (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done key_metrics — key-metrics
  • done key_metrics_ttm — key-metrics-ttm
  • done latest_financial_statements — latest-financial-statements (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done owner_earnings — owner-earnings
  • done ratios — ratios
  • done ratios_ttm — ratios-ttm
  • done revenue_geographic_segmentation — revenue-geographic-segmentation
  • done revenue_product_segmentation — revenue-product-segmentation

client.institutional_ownership — Form 13F institutional holdings, holders, and derived analytics.

Bucket 2 (Ultimate-gated), confirmed live: the pricing-audit prediction held — all 8 methods 402 on the free tier. Tests live in tests/ultimate/test_institutional_ownership.py only.

8 methods.

  • done institutional_ownership_dates — institutional-ownership/dates (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_extract — institutional-ownership/extract (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_extract_analytics_holder — institutional-ownership/extract-analytics/holder (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_holder_industry_breakdown — institutional-ownership/holder-industry-breakdown (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_holder_performance_summary — institutional-ownership/holder-performance-summary (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_industry_summary — institutional-ownership/industry-summary (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_latest — institutional-ownership/latest (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done institutional_ownership_symbol_positions_summary — institutional-ownership/symbol-positions-summary (works on Ultimate tier; 402 on free, Starter, and Premium)

client.indexes — Stock-market indexes, their quotes/charts, and their constituent lists.

7 methods.

Unit-tested and live-tested 2026-08-23. Only index_list is free-tier reachable — the other 6 own methods all 402 despite no Bucket 2 flag in REWRITE_ARCHITECTURE.md §3.5. 3 cross-listed from client.chart (historical_chart, historical_price_eod_full, historical_price_eod_light) already hold the identity invariant — verified via assert client.indexes.historical_chart is client.chart.historical_chart at write time.

  • done dowjones_constituent — dowjones-constituent (works on Premium tier; 402 on free and Starter tiers)
  • done historical_dowjones_constituent — historical-dowjones-constituent (works on Premium tier; 402 on free and Starter tiers)
  • done historical_nasdaq_constituent — historical-nasdaq-constituent (works on Premium tier; 402 on free and Starter tiers)
  • done historical_sp500_constituent — historical-sp500-constituent (works on Premium tier; 402 on free and Starter tiers)
  • done index_list — index-list
  • done nasdaq_constituent — nasdaq-constituent (works on Premium tier; 402 on free and Starter tiers)
  • done sp500_constituent — sp500-constituent (works on Premium tier; 402 on free and Starter tiers)

client.commodity — Commodity instruments: list, quotes, charts.

1 methods.

Unit-tested and live-tested 2026-08-23. quote/quote_short/ batch_commodity_quotes will cross-list here once client.quote is built; the 3 client.chart cross-listings are already wired.

  • done commodities_list — commodities-list

client.crypto — Cryptocurrency instruments: list, quotes, charts.

1 methods.

Unit-tested and live-tested 2026-08-23. Same cross-listing note as client.commodity above.

  • done cryptocurrency_list — cryptocurrency-list

client.fundraisers — Reg CF crowdfunding and Reg D/A equity offerings.

6 methods.

Unit-tested and live-tested 2026-08-23 — all 6 free-tier reachable.

  • done crowdfunding_offerings — crowdfunding-offerings
  • done crowdfunding_offerings_latest — crowdfunding-offerings-latest
  • done crowdfunding_offerings_search — crowdfunding-offerings-search
  • done fundraising — fundraising
  • done fundraising_latest — fundraising-latest
  • done fundraising_search — fundraising-search

client.forex — FX pairs: list, quotes, charts.

1 methods.

Unit-tested and live-tested 2026-08-23. Same cross-listing note as client.commodity above.

  • done forex_list — forex-list

client.insider_trades — Form 4 insider transactions, statistics, and beneficial-ownership acquisitions.

6 methods.

Unit-tested and live-tested 2026-08-23. Only insider_trading_latest and insider_trading_transaction_type are free-tier reachable — the other 4 all 402 despite no Bucket 2 flag in REWRITE_ARCHITECTURE.md §3.5.

  • done acquisition_of_beneficial_ownership — acquisition-of-beneficial-ownership (works on Starter tier; 402 on free tier)
  • done insider_trading_latest — insider-trading/latest
  • done insider_trading_reporting_name — insider-trading/reporting-name (works on Starter tier; 402 on free tier)
  • done insider_trading_search — insider-trading/search (works on Starter tier; 402 on free tier)
  • done insider_trading_statistics — insider-trading/statistics (works on Starter tier; 402 on free tier)
  • done insider_trading_transaction_type — insider-trading-transaction-type

client.market_performance — Sector/industry performance and P/E, snapshot and historical, plus market leaders.

11 methods.

Unit- and live-tested 2026-08-23 — all 11 free-tier reachable, no 402s.

  • done biggest_gainers — biggest-gainers
  • done biggest_losers — biggest-losers
  • done historical_industry_pe — historical-industry-pe
  • done historical_industry_performance — historical-industry-performance
  • done historical_sector_pe — historical-sector-pe
  • done historical_sector_performance — historical-sector-performance
  • done industry_pe_snapshot — industry-pe-snapshot
  • done industry_performance_snapshot — industry-performance-snapshot
  • done most_actives — most-actives
  • done sector_pe_snapshot — sector-pe-snapshot
  • done sector_performance_snapshot — sector-performance-snapshot

client.market_hours — Exchange trading sessions and holiday calendars.

3 methods.

Unit-tested and live-tested 2026-08-23 — all 3 free-tier reachable.

  • done all_exchange_market_hours — all-exchange-market-hours
  • done exchange_market_hours — exchange-market-hours
  • done holidays_by_exchange — holidays-by-exchange

client.technical_indicators — Computed technical indicator series.

9 methods.

Unit-tested and live-tested 2026-08-23 — all 9 402 on the free tier, the whole group gated, despite no Bucket 2 flag in REWRITE_ARCHITECTURE.md §3.5.

  • done technical_indicators_adx — technical-indicators/adx (works on Starter tier; 402 on free tier)
  • done technical_indicators_dema — technical-indicators/dema (works on Starter tier; 402 on free tier)
  • done technical_indicators_ema — technical-indicators/ema (works on Starter tier; 402 on free tier)
  • done technical_indicators_rsi — technical-indicators/rsi (works on Starter tier; 402 on free tier)
  • done technical_indicators_sma — technical-indicators/sma (works on Starter tier; 402 on free tier)
  • done technical_indicators_standarddeviation — technical-indicators/standarddeviation (works on Starter tier; 402 on free tier)
  • done technical_indicators_tema — technical-indicators/tema (works on Starter tier; 402 on free tier)
  • done technical_indicators_williams — technical-indicators/williams (works on Starter tier; 402 on free tier)
  • done technical_indicators_wma — technical-indicators/wma (works on Starter tier; 402 on free tier)

client.news — News, press releases, and FMP editorial articles.

10 methods.

Unit- and live-tested 2026-08-23. Only fmp_articles is free-tier reachable — the other 9 methods (the whole NewsArticleResult-shaped family: general/ press-releases/stock/crypto/forex, each with a "-latest" sibling) all 402'd, despite no Bucket 2 flag in REWRITE_ARCHITECTURE.md §3.5 → ultimate-pending.

  • done fmp_articles — fmp-articles
  • done news_crypto — news/crypto (works on Starter tier; 402 on free tier)
  • done news_crypto_latest — news/crypto-latest (works on Starter tier; 402 on free tier)
  • done news_forex — news/forex (works on Starter tier; 402 on free tier)
  • done news_forex_latest — news/forex-latest (works on Starter tier; 402 on free tier)
  • done news_general_latest — news/general-latest (works on Starter tier; 402 on free tier)
  • done news_press_releases — news/press-releases (works on Premium tier; 402 on free and Starter tiers)
  • done news_press_releases_latest — news/press-releases-latest (works on Premium tier; 402 on free and Starter tiers)
  • done news_stock — news/stock (works on Starter tier; 402 on free tier)
  • done news_stock_latest — news/stock-latest (works on Starter tier; 402 on free tier)

client.quote — Real-time and aftermarket quotes, single and batch.

16 methods.

Unit- and live-tested 2026-08-23. Completes the full §4.3 cross-listing table: quote/quote_short/batch_index_quotes/batch_commodity_quotes/ batch_crypto_quotes/batch_forex_quotes are wired into client.indexes/ commodity/crypto/forex in groups.py — all 10 cross-listings in the whole rewrite are wired, identity-asserted at write time. The 5 single-symbol methods (quote, quote_short, aftermarket_quote, aftermarket_trade, stock_price_change) are free-tier reachable; all 11 batch_* methods 402'd — a clean split along §7.6's own singular/plural distinction.

  • done aftermarket_quote — aftermarket-quote
  • done aftermarket_trade — aftermarket-trade
  • done batch_aftermarket_quote — batch-aftermarket-quote (works on Premium tier; 402 on free and Starter tiers)
  • done batch_aftermarket_trade — batch-aftermarket-trade (works on Premium tier; 402 on free and Starter tiers)
  • done batch_commodity_quotes — batch-commodity-quotes (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_crypto_quotes — batch-crypto-quotes (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_etf_quotes — batch-etf-quotes (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_exchange_quote — batch-exchange-quote (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_forex_quotes — batch-forex-quotes (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_index_quotes — batch-index-quotes (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_mutualfund_quotes — batch-mutualfund-quotes (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done batch_quote — batch-quote (works on Premium tier; 402 on free and Starter tiers)
  • done batch_quote_short — batch-quote-short (works on Premium tier; 402 on free and Starter tiers)
  • done quote — quote
  • done quote_short — quote-short
  • done stock_price_change — stock-price-change

client.sec_filings — SEC filing search, SEC company identity, and SIC industry classification.

12 methods.

Unit- and live-tested 2026-08-23. 10 of 12 are free-tier reachable; industry_classification_search and all_industry_classification 402'd → ultimate-pending. sec_profile's second parameter renders in FMP's own docs as cik-A, which reads like a table-rendering artifact rather than a real wire name — exposed here as plain cik; live-tested with symbol only (didn't spend an extra call probing the cik-A question), so that particular oddity is still unconfirmed either way.

  • done all_industry_classification — all-industry-classification (works on Starter tier; 402 on free tier)
  • done industry_classification_search — industry-classification-search (works on Starter tier; 402 on free tier)
  • done sec_filings_8k — sec-filings-8k
  • done sec_filings_company_search_cik — sec-filings-company-search/cik
  • done sec_filings_company_search_name — sec-filings-company-search/name
  • done sec_filings_company_search_symbol — sec-filings-company-search/symbol
  • done sec_filings_financials — sec-filings-financials
  • done sec_filings_search_cik — sec-filings-search/cik
  • done sec_filings_search_form_type — sec-filings-search/form-type
  • done sec_filings_search_symbol — sec-filings-search/symbol
  • done sec_profile — sec-profile
  • done standard_industrial_classification_list — standard-industrial-classification-list

client.earnings_transcript — Earnings-call transcripts and their availability metadata.

Confirmed Bucket 2 (Ultimate-gated), 2026-08-23: all 4 methods 402'd, matching the original pricing audit's prediction. earnings_transcript_list is cross-listed into client.directory (§4.3, itself fully gated) — consistent with this result.

4 methods.

  • done earning_call_transcript — earning-call-transcript (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done earning_call_transcript_dates — earning-call-transcript-dates (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done earning_call_transcript_latest — earning-call-transcript-latest (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done earnings_transcript_list — earnings-transcript-list (works on Ultimate tier; 402 on free, Starter, and Premium)

client.congress — U.S. Senate and House financial disclosures, trades, and member profiles.

12 methods.

Unit- and live-tested 2026-08-23. Only the parameterless house_latest/ senate_latest listings are free-tier reachable — every symbol/id/name-scoped lookup (10 methods) 402'd. §7.5's parameter-naming bug mirrored as documented: house_trades_by_id and senate_trades_by_id (and every other senateID-taking method, including the House ones) expose the Python parameter senate_id — the wire name really is senateID even on House endpoints, called out loudly in house_trades_by_id's own docstring so it doesn't read as our bug.

  • done house_latest — house-latest
  • done house_trades — house-trades (works on Starter tier; 402 on free tier)
  • done house_trades_by_id — house-trades-by-id (works on Starter tier; 402 on free tier)
  • done house_trades_by_name — house-trades-by-name (works on Starter tier; 402 on free tier)
  • done senate_latest — senate-latest
  • done senate_net_worth — senate-net-worth (works on Premium tier; 402 on free and Starter tiers)
  • done senate_net_worth_aggregated — senate-net-worth-aggregated (works on Premium tier; 402 on free and Starter tiers)
  • done senate_positions — senate-positions (works on Premium tier; 402 on free and Starter tiers)
  • done senate_profile — senate-profile (works on Premium tier; 402 on free and Starter tiers)
  • done senate_trades — senate-trades (works on Starter tier; 402 on free tier)
  • done senate_trades_by_id — senate-trades-by-id (works on Starter tier; 402 on free tier)
  • done senate_trades_by_name — senate-trades-by-name (works on Starter tier; 402 on free tier)

client.bulk — Whole-universe bulk downloads.

Confirmed Bucket 2 (Ultimate-gated), 2026-08-23: all 18 methods 402'd, matching the original pricing audit's prediction — the whole group, not a subset. 17 of 18 methods return every field as a JSON string, including semantically numeric/boolean fields — a real, documented quirk (see types/bulk.py's module docstring), not a transcription choice. profile_bulk is the lone exception (real JSON types) and reuses ProfileResult directly rather than duplicating it, since its example response is field-for-field identical to profile/profile_cik's.

18 methods.

  • done balance_sheet_statement_bulk — balance-sheet-statement-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done balance_sheet_statement_growth_bulk — balance-sheet-statement-growth-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done cash_flow_statement_bulk — cash-flow-statement-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done cash_flow_statement_growth_bulk — cash-flow-statement-growth-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done dcf_bulk — dcf-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done earnings_surprises_bulk — earnings-surprises-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done eod_bulk — eod-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done etf_holder_bulk — etf-holder-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done income_statement_bulk — income-statement-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done income_statement_growth_bulk — income-statement-growth-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done key_metrics_ttm_bulk — key-metrics-ttm-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done peers_bulk — peers-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done price_target_summary_bulk — price-target-summary-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done profile_bulk — profile-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done rating_bulk — rating-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done ratios_ttm_bulk — ratios-ttm-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done scores_bulk — scores-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)
  • done upgrades_downgrades_consensus_bulk — upgrades-downgrades-consensus-bulk (works on Ultimate tier; 402 on free, Starter, and Premium)

client.tipranks — TipRanks partner analyst data.

Update, 2026-08-24: re-tested through every standard FMP tier up to and including Ultimate — all 7 still 402. FMP's own error message says why: this needs a separate paid add-on ("TipRanks data boost"), not covered by any Free/Starter/Premium/Ultimate plan. Dax has decided not to buy it, so this group stays permanently untested (community add-on) — see the dated entries near the top of this file.

Original note, 2026-08-23: all 7 methods 402'd on the free tier, matching the workflow doc's treat-as-Ultimate-until-proven-otherwise call — the whole group. tipranks_pit_symbol and tipranks_pit_analyst share one response type (TipranksPointInTimeResult) — identical fields in both documented examples. The 3 summary methods (tipranks_symbol_summary/tipranks_analyst_summary/ tipranks_firm_summary) each get their own top-level type (different identifying field: symbol/expertUID/firmName) but share two small nested breakdown types (TipranksRecommendationBreakdown, TipranksAnalystActionBreakdown).

7 methods.

  • untested (community add-on) tipranks_analyst_summary — tipranks-analyst-summary (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)
  • untested (community add-on) tipranks_analysts — tipranks-analysts (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)
  • untested (community add-on) tipranks_firm_summary — tipranks-firm-summary (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)
  • untested (community add-on) tipranks_pit_analyst — tipranks-pit-analyst (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)
  • untested (community add-on) tipranks_pit_symbol — tipranks-pit-symbol (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)
  • untested (community add-on) tipranks_search — tipranks-search (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)
  • untested (community add-on) tipranks_symbol_summary — tipranks-symbol-summary (402 on every plan tier through Ultimate — requires FMP's separate TipRanks add-on; Dax has decided not to buy it, so this stays untested)