for agents.
Connect your agent to pre-computed market context that improves reasoning and reduces token usage.
No credit card required
$ curl "https://api.tickerdb.com/v1/summary/NVDA" -H "Authorization: Bearer tdb_your_api_key"
{
"ticker": "NVDA",
"trend": {
"direction": "strong_uptrend",
"duration_days": 34,
"ma_alignment": "aligned_bullish",
"volume_confirmation": "confirmed"
},
"momentum": {
"rsi_zone": "overbought",
"macd_state": "expanding_positive",
"divergence_detected": true
},
"extremes": {
"condition": "overbought",
"condition_rarity": "rare",
"condition_percentile": 7.3
},
"resistance_level": {
"status": "approaching",
"distance_band": "very_close",
"touch_count": 4
},
"fundamentals": {
"valuation_zone": "undervalued",
"growth_zone": "high_growth",
"analyst_consensus": "strong_buy"
}
}
Improve reasoning with actionable context. #
Before your agent can reason about an asset, it needs proper market context. TickerDB computes the context for you.
[
{
"date": "2024-01-15",
"open": 182.16,
"high": 184.26,
"low": 180.93,
"close": 183.63,
"volume": 65234100
},
// ... 200 more rows
]
- LLM has to compute raw data
- Uses more tokens
- Not designed for agents
- Inconsistent reasoning
{
"trend": {
"direction": "uptrend",
"ma_alignment": "aligned_bullish"
},
"momentum": {
"rsi_zone": "neutral_high"
},
"volatility": {
"regime": "normal"
}
}
- LLM reads categorical bands it understands
- Lower token usage
- Designed for agents
- Consistent reasoning
The context your agent needs. #
Pre-computed
Our data is computed after market close and cached. Zero latency on your request - no indicator math, no delays.
Categorical vocabulary
Responses use terms like oversold
, uptrend
, and deeply_overvalued
- the same vocabulary LLMs use to reason about markets.
Fewer wasted tokens
A full asset summary is a fraction of the tokens you'd need to pass raw OHLCV. Your model gets more context, not more noise.
Multi-source intelligence
Technical, fundamental, and analyst data in one response. No stitching together multiple providers.
Per-asset behavioral context
Historical streaks, medians, and percentiles specific to each asset. Your agent knows what's unusual for this ticker.
5 years of queryable history
Every categorical field, every day, for 5 years. Just connect and query.
Built for workflows agents struggle with. #
TickerDB is strongest when your agent needs actionable market context, historical precedent, or watchlist diffs instead of raw price bars.
Watchlist monitoring
Track saved assets and pull only what changed with /v1/watchlist/changes
. Ideal for daily alerts, portfolio briefings, and autonomous monitoring.
Market scanning
Use /v1/search
to find oversold names, strong uptrends, rare conditions, or valuation mismatches across the full universe.
Historical precedent
Query /v1/summary
with field
and band
to see when a setup last appeared and what happened after.
Track state changes effortlessly. #
TickerDB monitors your watchlist. When something meaningful changes, a structured diff is generated. Pull it on demand or get it pushed via webhooks.
/v1/watchlist/changes
{
"timeframe": "daily",
"run_date": "2026-03-28",
"changes": {
"AAPL": [
{
"field": "rsi_zone",
"from": "neutral",
"to": "oversold"
},
{
"field": "divergence_detected",
"from": false,
"to": true
}
],
"TSLA": [
{
"field": "macd_state",
"from": "contracting_negative",
"to": "expanding_positive"
}
],
"BTCUSD": [
{
"field": "squeeze_active",
"from": false,
"to": true
}
]
},
"tickers_checked": 12,
"tickers_changed": 3
}
{
"timeframe": "daily",
"run_date": "2026-03-28",
"changes": {
"AAPL": [
{
"field": "rsi_zone",
"from": "neutral",
"to": "oversold"
},
{
"field": "divergence_detected",
"from": false,
"to": true
}
]
},
"tickers_checked": 12,
"tickers_changed": 1
}
Add tickers to your watchlist
Track the assets you care about. Stocks, crypto, or both.
TickerDB computes daily diffs
After each pipeline run, every tracked field is compared against the prior day. Only assets with at least one change are included.
Your agent reads only what changed
No full snapshots to diff yourself. No wasted tokens on data that hasn't moved. Just the fields that shifted, with from
and to
values your agent can act on.
Integrate in minutes. #
Make your first call in minutes. Drop in our SDK or make direct HTTP calls.
$ curl "https://api.tickerdb.com/v1/summary/AAPL" \
-H "Authorization: Bearer YOUR_API_KEY"
$ curl -G "https://api.tickerdb.com/v1/search" \
--data-urlencode 'filters=[{"field":"momentum_rsi_zone","op":"eq","value":"oversold"}]' \
-H "Authorization: Bearer YOUR_API_KEY"
$ curl https://api.tickerdb.com/v1/watchlist/changes \
-H "Authorization: Bearer YOUR_API_KEY"
python
import requests
res = requests.get(
"https://api.tickerdb.com/v1/summary/AAPL",
headers={"Authorization": "Bearer YOUR_API_KEY"}
)
data = res.json()
prompt = f"""
Analyze AAPL based on this market data:
{data}
Is this a good entry point?
"""
js
import { TickerDB } from 'tickerdb';
// One call. Full market context.
const client = new TickerDB({ apiKey: 'YOUR_API_KEY' });
const { data } = await client.summary('AAPL');
// Hand it directly to your LLM
const prompt = `Analyze AAPL based on this market data:
$${JSON.stringify(data)$}
Is this a good entry point?`;
// Ready for your LLM
import "context"
import "github.com/tickerdb/tickerdb-go"
// One call. Full market context.
client := tickerdb.NewClient("YOUR_API_KEY")
resp, _ := client.Summary(context.Background(), "AAPL", nil)
// Raw JSON plus rate limits
fmt.Println(string(resp.Data))
fmt.Println(resp.RateLimits.RequestsRemaining)
// claude_desktop_config.json
{
"mcpServers": {
"tickerdb": {
"command": "npx",
"args": ["tickerdb-mcp"],
"env": {
"TICKERDB_KEY": "YOUR_API_KEY"
}
}
}
}
Once connected, Claude can call get_summary
, get_search
, get_watchlist_changes
, and more - directly from the chat. all MCP tools →
$ clawhub install tickerdb
"How's AAPL looking? Give me the full summary."
"Add NVDA, AAPL, and BTCUSD to my watchlist and flag anything that changed."
"When was NVDA last deep_oversold, and what happened after?"
One install. Your agent gets summaries, search, watchlists, and schema - no config needed. OpenClaw integration guide →
Compatible with everything. #
MCP, OpenClaw, SDKs or plain HTTP. If it can make a GET request, it works with TickerDB.
MCP Server
Plug TickerDB into Claude Desktop or any MCP client. Your AI assistant pulls EOD market context from the chat.
Any agent framework
LangChain, LlamaIndex, AutoGen, CrewAI - it's just HTTP. If it can make a GET request, it works.
Python, Node.js & Go SDKs
Official SDKs with typed responses. Or just use fetch
- the API is simple enough.
One database. Infinite ways to query. #
Over 140 queryable fields across 10,000+ assets.