TickerDB
OfficialServer Configuration
Describes the environment variables required to run the server.
| Name | Required | Description | Default |
|---|---|---|---|
| TICKERDB_API_KEY | No | Optional. Your TickerDB API key (tdb_...). Without it the server reads ./.env or the config saved by device login at tickerdb.com/setup. |
Instructions
Guidance the server publishes about itself, which clients place ahead of the tool catalog so the model reads it before choosing anything.
TickerDB: point-in-time market data for stocks, ETFs and crypto, as named bands (momentum_rsi_zone = "oversold") plus raw values. Look up field names with get_schema (search="rsi zone", level=market) instead of guessing. null means unknown, not zero or low. Every band also has _prev, _periods, _flips_recent_count, _flips_lookback and _stability, usable as fields and filters; _periods = 1 means it changed on that row. Use query for any cross-section or history: filter and sort on the server rather than pulling rows and filtering yourself. For multi-year research prefer timeframe=weekly. For large pulls pass store_as= plus fetch_all=true: rows go into a SQLite table, you get a short summary, and query_data reads it with SQL. The user's watchlist (one per account): add_watchlist_tickers and remove_watchlist_tickers change it; get_watchlist_changes (call_tool) says what changed since the last check, for a daily check. Other endpoints (changes, watchlist, insiders, holdings, funds, indexes, TickerDB's own trading models, status, account): search_tools, then call_tool. A row dated D is known on D+1: act at the next session's open.
Capabilities
Features and capabilities supported by this server
Protocol revision2025-11-25
| Capability | Details |
|---|---|
| tools | {
"listChanged": true
} |
Tools
Functions exposed to the LLM to take actions
| Name | Description |
|---|---|
| get_schemaA | Look up field names, descriptions, units and band values. Search rather than pulling everything (~230 KB): search="bull flag", category=patterns, level=market. Names are []; bands end in a category word (_zone, _state, _trend), raw numbers in a unit (_pct, _ratio, _usd). Every band (tracked: true) also has _prev, _periods, _flips_recent_count, _flips_lookback and _stability, usable as fields and filters; lookups omit those rows unless markers=true. Over 25 matches lists names only: search again by exact name for a description. |
| queryA | Any cross-section or history (one ticker's full snapshot: get_summary). Give fields and filters (names from get_schema), then EITHER date (default latest) OR start+end. One date with filters, tickers, sort_by or limit: a ranked screen. One date with only fields: every asset. A range: every row in it, weekly by default past ~6 months; market_/macro_ filters keep only dates in that regime, other filters apply after the rows are fetched, so narrow with tickers, member_of or asset_class. Big pulls: store_as + fetch_all, then query_data. |
| get_summaryA | One ticker's state: latest snapshot, one date, a start/end series (up to 50 rows; for a longer range add sample=even or use query), or band events. For events give field and optionally band: each says when the band was entered, what it left and how long that had lasted; stats=true returns how often each band was entered. fields trims to sections, paths or field names, e.g. trend, momentum.rsi_zone, fundamentals_pe_ratio. Field names are the get_schema names everywhere (no aliases). Market regimes are in get_market. |
| get_marketA | Market-level rows, one per date, point in time: US equity market regime (market_*) and economy (macro_*) bands and values, e.g. market_regime_zone, macro_yield_curve_zone. changes=true returns only dates where a band changed. Field list: get_schema level=market. To find tickers under a regime, put market_/macro_ filters into query. |
| get_ohlcvA | End-of-day OHLCV candles for one ticker, daily or weekly (weeks dated by Sunday). Paged: pass next_cursor back as cursor, or use store_as with fetch_all for the whole range in one call. |
| add_watchlist_tickersA | Add tickers to the user's TickerDB watchlist (one per account, shared with the website's Track buttons and Claude/ChatGPT watchlist links). Idempotent: already-present tickers come back in already; unknown ones in unknown. No ticker limit; free. Then get_watchlist_changes (call_tool) reports what changed for them since the last check: score, upside and downside potential, earnings, big price moves, trend changes, RSI oversold, volume spikes, insider trades, analyst changes, serious 8-K filings and dividend cuts. |
| remove_watchlist_tickersA | Remove tickers from the user's TickerDB watchlist. Idempotent: tickers not on it come back in missing. Free. |
| search_toolsA | Find the other TickerDB endpoints: changes, the watchlist and what changed for it, insiders, insider trades and holdings, funds, fund holdings and holders, indexes, index members, TickerDB's trading models (list, detail, history, holdings, trades; Pro and Commercial plans, except a model's newest holdings, on every plan), status, account, and the raw search and universe routes. Returns each match's name, purpose and params (* = required). Empty query lists all. Run one with call_tool. |
| call_toolA | Run an endpoint found with search_tools: its name plus a params object. store_as, append and fetch_all work as on the other data tools. |
| query_dataA | Read-only SQL (SQLite) over tables saved with store_as: SELECT or WITH, plus SHOW TABLES and DESCRIBE . Returns at most 200 rows; output_path writes the full result to a CSV file for bulk work. Nested fields are flattened with _ (trend.direction -> trend_direction); booleans are 1/0. Tables live in memory for this server process. |
Prompts
Interactive templates invoked by user choice
| Name | Description |
|---|---|
No prompts | |
Resources
Contextual data attached and managed by the client
| Name | Description |
|---|---|
No resources | |
TDQS
Scored across 10 tools
Most tools occupy distinct niches (OHLCV candles, market-level bands, one-ticker summary, cross-sectional screens, SQL over saved tables), and descriptions cross-reference each other to route the agent correctly. The main ambiguity is between query and query_data, whose names suggest a variant relationship but whose purposes (screening vs. SQL over stored tables) are very different, and the search_tools/call_tool meta-pair adds an extra indirection layer an agent must understand.
Names are all snake_case with a clear verb_noun shape (get_market, get_ohlcv, get_schema, get_summary, add_watchlist_tickers, remove_watchlist_tickers, search_tools, call_tool), and the add/remove pair is symmetric. The deviations are bare 'query' and the confusable 'query_data', which breaks the get_/verb prefix pattern and collides semantically with 'query'.
Ten tools is well within a healthy range, and each earns its place: watchlist mutations, market/OHLCV/summary retrieval, schema lookup, screening, SQL, and a discovery layer for the long tail of endpoints. The design deliberately pushes the remaining endpoints behind search_tools/call_tool instead of bloating the surface, which is exactly the right scoping.
Core workflows are covered end to end: add/remove watchlist tickers, read what changed, pull prices, screen, inspect schema, and persist results for SQL analysis, with the rest of the domain (insiders, funds, indexes, models, account/status) reachable via search_tools + call_tool. The minor gap is that reading the plain watchlist or its changes requires the indirect call_tool path rather than a first-class tool, which is a small discoverability cost rather than a dead end.