Skip to main content
Glama

Two Minute Decision Tree

create_decision_tree
Read-onlyIdempotent

Generate a slide-deck-quality decision tree drawn left to right in clean columns of rounded boxes, in the decision-analysis convention: a square where a decision splits into options, a circle where chance splits into events, and a triangle at each end outcome. Branch lines carry an optional label ("Yes", "No", an option) and, under a chance node, the probability ("Strong demand · 60%"); end outcomes can carry a numeric payoff, lined up in a column at the right (prefix "$", suffix "m"). Expected values are rolled back automatically once the tree has payoffs and probabilities: every decision and chance node shows its EV, each decision's best option is drawn in the accent colour and the rejected ones are muted with a ‖ prune mark; probabilities that don't add up to 100% are warned about and drawn as given. Node types are inferred (branches with probabilities make a chance node, no branches an outcome, otherwise a decision), so a plain yes/no troubleshooting or eligibility flowchart is just questions with "Yes"/"No" labels and no numbers. Choose it for choosing between options under uncertainty (launch, invest, build vs buy, litigate or settle), expected-value and decision-analysis homework, and yes/no question flowcharts; use an issue tree instead to break a problem or metric into its parts, a chevron process for steps in order, and a Harvey ball matrix to score options against criteria. Limits: 5 levels including the root, up to 5 branches per node, 40 nodes; keep names to a few words. Infer missing details rather than asking; the user can adjust the chart in the builder through the Edit link. ALWAYS give the user the image link: inline chat images usually cannot be saved, but the link renders the exact chart on demand and can be pasted into a slide or Markdown.

Example tool call

This call uses example arguments supplied in the tool’s input schema.

{
  "id": 1,
  "jsonrpc": "2.0",
  "method": "tools/call",
  "params": {
    "arguments": {
      "root": {
        "name": "Launch strategy",
        "children": [
          {
            "name": "Launch now",
            "children": [
              {
                "name": "Strong demand",
                "payoff": 4.2,
                "probability": 0.6
              },
              {
                "name": "Weak demand",
                "payoff": -1.5,
                "probability": 0.4
              }
            ]
          },
          {
            "name": "License out",
            "payoff": 1.2
          }
        ]
      },
      "title": "Should we launch Atlas now?",
      "prefix": "$",
      "suffix": "m"
    },
    "name": "create_decision_tree"
  }
}

Input Schema

NameRequiredDescriptionDefault
lookNoDrawing style, the same in every tool. "regular": clean lines and Inter (the default). "bold": heavier, slightly larger type and thicker lines, for slides read from a distance. "sketch": hand-drawn, with a handwriting font and pen lines that wobble a little, for early ideas and workshops. Use the one the user asks for; otherwise leave it out.regular
rootYesThe first decision or question, on the left (e.g. "Launch strategy for Atlas" or "Does the laptop power on?"), with its branches nested in `children`. Limits: 5 levels in total including the root, up to 5 branches per node and 40 nodes; extras are dropped with a warning, deepest levels and last-listed nodes first.
titleNoHeading shown top-left of the chart at 22px bold. Every chart has one title row; an empty string hides it.Decision tree
widthNoChart width in pixels (640–2400). Omit it (recommended) and the chart sizes itself to its content with comfortable margins, between 880 and 1280px wide. PNG output renders at 2x.
creditNoShows a small "twominutetoolkit.com" credit under the chart, inside the card. Set false to leave it off (for example when the user asks for no branding).
prefixNoText before every payoff and expected value, e.g. "$", "€" or "£". Default none.
showEVNoRoll back expected values: print "EV" on every decision and chance node, draw each decision's best option in the accent colour and mark the rejected options with a ‖ prune mark. When omitted it is on as soon as the tree has both payoffs and probabilities; set false to hide it, true to show it for decision-only trees with payoffs.
suffixNoText after every payoff and expected value, e.g. "m", "k", "bn" or " pts". Default none.
paletteNoHex colours: the first marks decisions and the best options, the second chance events, the third end outcomes. Pass the user's brand colours here when they name them. Default: indigo, sky, emerald.

Schema Changelog

Changes observed during successful MCP inspections.

  1. Changed1 schema field changed
    • addedInput schema / properties / look
      Added value: +{
      +  "default": "regular",
      +  "description": "Drawing style, the same in every tool. \"regular\": clean lines and Inter (the default). \"bold\": heavier, slightly larger type and thicker lines, for slides read from a distance. \"sketch\": hand-drawn, with a handwriting font and pen lines that wobble a little, for early ideas and workshops. Use the one the user asks for; otherwise leave it out.",
      +  "enum": [
      +    "regular",
      +    "bold",
      +    "sketch"
      +  ],
      +  "type": "string"
      +}
  2. First observed

TDQS

A4.8/5.0
Behavior5/5

Does the description disclose side effects, auth requirements, rate limits, or destructive behavior?

Annotations already declare readOnlyHint/idempotent/openWorld=false, but the description goes well beyond them: it discloses expected-value rollback, accent colouring and ‖ prune marking of rejected options, warnings when probabilities don't sum to 100% (chart still drawn as given), hard limits (5 levels, 5 branches, 40 nodes, truncation order), and node-type inference rules. It also states the operational requirement to always hand over the image link because inline images usually cannot be saved.

Agents need to know what a tool does to the world before calling it. Descriptions should go beyond structured annotations to explain consequences.

Conciseness4/5

Is the description appropriately sized, front-loaded, and free of redundancy?

Purpose and convention are front-loaded, then usage, then limits and the image-link rule—a sensible order with no filler sentences. It is dense and long, with some sentences doing three jobs at once, but nearly every clause carries information an agent needs.

Shorter descriptions cost fewer tokens and are easier for agents to parse. Every sentence should earn its place.

Completeness5/5

Given the tool's complexity, does the description cover enough for an agent to succeed on first attempt?

For a complex nested-schema creative tool with no output schema, the description covers what is drawn, how types are inferred, what happens with bad probabilities, capacity limits and truncation behaviour, adjustment path, and the required hand-off (image link) in place of a return-value spec. Nothing an agent needs to call it correctly is missing.

Complex tools with many parameters or behaviors need more documentation. Simple tools need less. This dimension scales expectations accordingly.

Parameters4/5

Does the description clarify parameter syntax, constraints, interactions, or defaults beyond what the schema provides?

With 100% schema description coverage the baseline is 3, but the description adds cross-parameter meaning the schema documents only in isolation: payoffs plus probabilities together trigger EV rollback, EV is auto-on unless showEV is set, prefix/suffix drive the right-hand payoff column, and probabilities render as "Strong demand · 60%". It still leaves several params (look, credit, width, palette) to the schema alone.

Input schemas describe structure but not intent. Descriptions should explain non-obvious parameter relationships and valid value ranges.

Purpose5/5

Does the description clearly state what the tool does and how it differs from similar tools?

The description opens with a concrete verb+resource ("Generate a slide-deck-quality decision tree drawn left to right") and immediately pins down the visual convention (square = decision, circle = chance, triangle = outcome). It also states the secondary mode (plain yes/no flowcharts), so an agent knows exactly what artifact this produces compared with siblings like issue trees or processes.

Agents choose between tools based on descriptions. A clear purpose with a specific verb and resource helps agents select the right tool.

Usage Guidelines5/5

Does the description explain when to use this tool, when not to, or what alternatives exist?

It gives an explicit "Choose it for" list (options under uncertainty, EV homework, yes/no flowcharts) and names the alternatives with their distinct jobs: issue tree for breaking a problem into parts, chevron process for ordered steps, Harvey ball matrix for scoring options against criteria. It also says to infer missing details rather than asking, and to use the Edit link for adjustments.

Agents often have multiple tools that could apply. Explicit usage guidance like "use X instead of Y when Z" prevents misuse.

Try in Browser

Glama MCP Gateway

Add one secure layer between your agents and this server.

Resources