Draw a Clarifo chart of a result you already fetched. Three outputs:

• output="python" — ChatGPT and any host with a python / code tool. Returns a ready-to-run matplotlib script with every number, label, colour and the source line embedded server-side. Run it unchanged in your python tool; it ends with plt.show(), which makes the chart appear in your answer. Do not edit the numbers, do not save the figure to a file and never write a Markdown image link (![...](sandbox:...)) — it renders as a broken image.

• output="image" (default) — a PNG image block, for hosts that show MCP images to the user inline (Claude). ChatGPT does NOT show MCP images to the user: use output="python" there.

• output="chart_spec" — the validated chart as JSON (periods, one series per metric with its unit, formula and figure basis, per-period citations, quality warnings, a text alternative). For a host that draws charts itself: plot these values exactly and print the warnings; it does not guarantee the host shows an interactive chart.

All three are built from one spec, so they always show the same metrics, periods and figures.

Call it only when a chart earns its place (a trend over 3+ periods, a peer comparison over time, a ranking of 4+ names, a sensitivity grid); a single figure or a two-row comparison needs a sentence or a table. For several metrics of one company pass metrics (e.g. ["revenue", "operating_margin"]: revenue bars on the left axis, the margin as a line on a labelled right axis); a peer comparison takes one metric per call. Only the named metrics are drawn — an unknown metric, or one the result has no values for, is an error, never a substitute. Years with no value for any charted metric are left off the axis; interior gaps stay gaps.

Pass the name of the tool you called and the same arguments you passed it. The data is re-fetched and the chart is computed server-side from Clarifo's stored data — never pass figures yourself, there is no parameter for them. For get_stock_price pass the history-mode arguments (days or a date range); without them that tool performs a live latest-close web lookup.

Chartable tools: compare_companies (multi-year: one line per company for metric, calendar-year aligned; single period: KPI table), forecast_financials, get_company_financials, ownership (all three modes), get_revenue_geography, get_stock_price (history mode, i.e. with days or a date range), get_valuation_history, screen_companies, screen_metric_trend, screen_quality_growers, value_company.

Args: source_tool: Name of the tool whose result to chart. source_args: The arguments you passed to that tool, verbatim. metric: Which metric to chart where the source carries several (compare_companies: revenue (default), operating_margin, net_income, gross_margin, free_cash_flow, ...). get_company_financials: revenue_operating_income draws revenue and operating income in one stacked bar per year (bar total = revenue, base segment = operating income, margin row underneath; a loss hangs below zero). Annual data only. Single-company metrics: revenue, gross_profit, operating_income, ebit, ebitda, net_income, operating_cash_flow, capital_expenditures, free_cash_flow, total_assets, total_equity, eps_diluted (reported) and gross_margin, operating_margin, ebit_margin, ebitda_margin, net_margin, fcf_margin (computed: amount / revenue). output: "python" (ChatGPT / code tools), "image" (PNG) or "chart_spec" (JSON for a host that draws natively). metrics: Several metrics in one chart (get_company_financials); metric alone keeps working. chart_type: "auto" (default: bars for up to 5 annual periods, a line for longer series; an amount with a margin is a bar_line_combo for up to 8 periods), "line", "bar", "grouped_bar", "horizontal_bar" or "bar_line_combo" (one company: amounts as bars, the % metric as a line on its own right axis). Applies to get_company_financials and compare_companies (combo: single company only).

Returns: python: the script plus run instructions and the source line to print under the chart. image: a one-line description plus the PNG. chart_spec: the chart as data. All three carry quality_warnings to print under the chart when present.