Market Briefing

Market Briefing

Fetch quotes for one or more symbols, speak a tight summary, and render a stocks_ticker (plus an optional chart_3d for trend).

Steps

1. Resolve symbols. From the request ("AAPL and Tesla") or the user's saved

watchlist; default to a small index set if none given.

1. Call get_stocks{symbols}. Returns data.quotes ({symbol, price,

change_pct, currency}) and a stocks_ticker directive.

1. Summarize movers in agent.speech (biggest up/down).
2. Optional trend: for "show the trend" build a chart_3d (line) from a

series of closes.

1. Helper: to assemble ticker props from raw SYMBOL:price[:change_pct]

pairs deterministically, use the bundled script: python3 scripts/build_ticker.py AAPL:214.5:1.2 TSLA:178.3:-0.8.

Output

stocks_ticker (show_stocks, props per registry.json):

{ "widget_type": "stocks_ticker",
  "transform": { "anchor": "head", "billboard": true },
  "props": { "title": "Watchlist",
             "symbols": [ { "symbol": "AAPL", "price": 214.5, "change_pct": 1.2, "currency": "USD" },
                          { "symbol": "TSLA", "price": 178.3, "change_pct": -0.8, "currency": "USD" } ],
             "scroll": true },
  "interactions": ["grab","tap","resize","drag"] }

Trend (show_chart → chart_3d):

{ "widget_type": "chart_3d",
  "props": { "chart_type": "line", "title": "AAPL — 5d",
             "labels": ["Mon","Tue","Wed","Thu","Fri"],
             "series": [ { "name": "Close", "values": [210,212,209,213,214.5], "color": "#4FC3F7" } ],
             "y_axis_label": "USD" } }

agent.speech:

{ "text": "Markets: AAPL is up 1.2% at $214.50, Tesla down 0.8% at $178.30.", "final": true }

Bundled resources

• scripts/build_ticker.py — pure-stdlib helper that turns CLI SYMBOL:price

pairs into validated stocks_ticker props JSON (mirrors the widget schema).

Edge cases

• Unknown symbol → omit it and mention it wasn't found; don't guess a price.
• Markets closed / stale → label values as last close.
• Crypto / FX → same ticker schema; set currency appropriately.
• Offline → values are deterministic mocks; say so for "live price" asks.
• Large watchlist (>6) → ticker caps at a handful; suggest a data_table

for the full list via stage-agent present-data.

Source & license

This open-source skill is cataloged on AgentStack and links to its original source — we do not rehost the code.

• Author: sumitaich1998
• Source: sumitaich1998/jarvisvr
• License: MIT

Install and usage instructions live in the source repository linked above.