I pull a live gold signal and a macro snapshot every morning from my own data API. On August 19, 2026 the two did not tell the same story, and that gap is more useful than either number on its own.
what the signal said
The XAUUSD signal came back SHORT. Entry 4408.4, stop loss 4441.17, take profit 1 at 4359.24. Confidence 0.72. The engine is a live H1 moving average crossover, MA20 crossing under MA50, with a stated horizon of one to four hours. Live price was 4408.4 at the time of the read.
That is a short term mean reversion read. It is not a view on where gold closes next week.
what the macro tape showed
The macro snapshot for the same day was mixed but mostly gold friendly:
- DXY 99.635, flat at minus 0.005 percent
- US 10 year yield 4.706, down 0.38 percent
- VIX 15.84, up 4.28 percent, risk off leaning
- SPX 7691.76, down 0.69 percent
- Silver 63.39, down 4.13 percent
- BTC 64388.55, down 0.18 percent
Falling real yields and a soft dollar usually support gold on a multi day horizon. Silver dumping more than 4 percent in a session is the one loud warning sign in the bundle.
why I care about the gap
The signal is a one to four hour momentum model. The macro read is a slower, cross asset backdrop. They answer different questions. A short term pullback call and a constructive bigger picture setup can both be right at the same time, just on different clocks.
What I would not do is treat the 0.72 confidence as strong conviction. It is a moderate read, and the silver move says the tape is not calm.
pull it yourself
The data is served as plain JSON over an x402 endpoint. A quick read looks like this:
curl -s https://agentdatum.com/api/v1/gold/signal
curl -s https://agentdatum.com/api/v1/gold/macro
The full machine readable catalog of 284 resources, including these two endpoints, is here:
https://agentdatum.com/.well-known/ai-catalog.json
honest limits
Two caveats I want to be straight about. First, the signal is a single timeframe. It has no opinion on the daily or weekly trend, and it resets fast. Second, the macro bundle is a snapshot, not a forecast. DXY and the 10 year can flip within a session, and the gold friendly reading can vanish before the short term signal plays out. I use both, but I size positions for the shorter clock and I do not pretend the macro read confirms the trade.
If you want to build on the same endpoints, start from the catalog above. It lists every field, the update cadence, and the payment scheme in one file.
This post was written with AI assistance for drafting and editing.
Top comments (0)