# Send Decision Trees to your agent Hand the extracted package to your coding agent with a concrete install brief instead of figuring it out manually. ## Fast path - Download the package from Yavira. - Extract it into a folder your agent can access. - Paste one of the prompts below and point your agent at the extracted folder. ## Suggested prompts ### New install ```text I downloaded a skill package from Yavira. Read SKILL.md from the extracted folder and install it by following the included instructions. Then review README.md for any prerequisites, environment setup, or post-install checks. Tell me what you changed and call out any manual steps you could not complete. ``` ### Upgrade existing ```text I downloaded an updated skill package from Yavira. Read SKILL.md from the extracted folder, compare it with my current installation, and upgrade it while preserving any custom configuration unless the package docs explicitly say otherwise. Then review README.md for any prerequisites, environment setup, or post-install checks. Summarize what changed and any follow-up checks I should run. ``` ## Machine-readable fields ```json { "schemaVersion": "1.0", "item": { "slug": "decision-trees", "name": "Decision Trees", "source": "tencent", "type": "skill", "category": "AI 智能", "sourceUrl": "https://clawhub.ai/evgyur/decision-trees", "canonicalUrl": "https://clawhub.ai/evgyur/decision-trees", "targetPlatform": "OpenClaw" }, "install": { "downloadUrl": "/downloads/decision-trees", "sourceDownloadUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=decision-trees", "sourcePlatform": "tencent", "targetPlatform": "OpenClaw", "packageFormat": "ZIP package", "primaryDoc": "SKILL.md", "includedAssets": [ "README.md", "SKILL.md", "scripts/decision_tree.py" ], "downloadMode": "redirect", "sourceHealth": { "source": "tencent", "slug": "decision-trees", "status": "healthy", "reason": "direct_download_ok", "recommendedAction": "download", "checkedAt": "2026-05-01T01:12:54.738Z", "expiresAt": "2026-05-08T01:12:54.738Z", "httpStatus": 200, "finalUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=decision-trees", "contentType": "application/zip", "probeMethod": "head", "details": { "probeUrl": "https://wry-manatee-359.convex.site/api/v1/download?slug=decision-trees", "contentDisposition": "attachment; filename=\"decision-trees-1.0.1.zip\"", "redirectLocation": null, "bodySnippet": null, "slug": "decision-trees" }, "scope": "item", "summary": "Item download looks usable.", "detail": "Yavira can redirect you to the upstream package for this item.", "primaryActionLabel": "Download for OpenClaw", "primaryActionHref": "/downloads/decision-trees" }, "validation": { "installChecklist": [ "Use the Yavira download entry.", "Review SKILL.md after the package is downloaded.", "Confirm the extracted package contains the expected setup assets." ], "postInstallChecks": [ "Confirm the extracted package includes the expected docs or setup files.", "Validate the skill or prompts are available in your target agent workspace.", "Capture any manual follow-up steps the agent could not complete." ] } }, "links": { "detailUrl": "https://openagent3.xyz/skills/decision-trees", "downloadUrl": "https://openagent3.xyz/downloads/decision-trees", "agentUrl": "https://openagent3.xyz/skills/decision-trees/agent", "manifestUrl": "https://openagent3.xyz/skills/decision-trees/agent.json", "briefUrl": "https://openagent3.xyz/skills/decision-trees/agent.md" } } ``` ## Documentation ### Decision Trees — Structured Decision-Making Decision tree analysis: a visual tool for making decisions with probabilities and expected value. ### When to Use ✅ Good for: Business decisions (investments, hiring, product launches) Personal choices (career, relocation, purchases) Trading & investing (position sizing, entry/exit) Operational decisions (expansion, outsourcing) Any situation with measurable consequences ❌ Not suitable for: Decisions with true uncertainty (black swans) Fast tactical choices Purely emotional/ethical questions ### Method Decision tree = tree-like structure where: Decision nodes (squares) — your actions Chance nodes (circles) — random events End nodes (triangles) — final outcomes Process: Define options — all possible actions Define outcomes — what can happen after each action Estimate probabilities — how likely is each outcome (0-100%) Estimate values — utility/reward for each outcome (money, points, utility units) Calculate EV — expected value = Σ (probability × value) Choose — option with highest EV ### Formula EV = Σ (probability_i × value_i) Example: Outcome A: 70% probability, +$100 → 0.7 × 100 = $70 Outcome B: 30% probability, -$50 → 0.3 × (-50) = -$15 EV = $70 + (-$15) = $55 ### Classic Example (from Wikipedia) Decision: Go to party or stay home? ### Estimates: Party: +9 utility (fun) Home: +3 utility (comfort) Carrying jacket unnecessarily: -2 utility Being cold: -10 utility Probability cold: 70% Probability warm: 30% ### Tree: Decision ├─ Go to party │ ├─ Take jacket │ │ ├─ Cold (70%) → 9 utility (party) │ │ └─ Warm (30%) → 9 - 2 = 7 utility (carried unnecessarily) │ │ EV = 0.7 × 9 + 0.3 × 7 = 8.4 │ └─ Don't take jacket │ ├─ Cold (70%) → 9 - 10 = -1 utility (froze) │ └─ Warm (30%) → 9 utility (perfect) │ EV = 0.7 × (-1) + 0.3 × 9 = 2.0 └─ Stay home └─ EV = 3.0 (always) Conclusion: Go and take jacket (EV = 8.4) > stay home (EV = 3.0) > go without jacket (EV = 2.0) ### Business Example Decision: Launch new product? ### Estimates: Success probability: 40% Failure probability: 60% Profit if success: $500K Loss if failure: $200K Don't launch: $0 ### Tree: Launch product ├─ Success (40%) → +$500K └─ Failure (60%) → -$200K EV = (0.4 × 500K) + (0.6 × -200K) = 200K - 120K = +$80K Don't launch └─ EV = $0 Conclusion: Launch (EV = +$80K) is better than not launching ($0). ### Trading Example Decision: Enter position or wait? ### Estimates: Probability of rise: 60% Probability of fall: 40% Position size: $1000 Target: +10% ($100 profit) Stop-loss: -5% ($50 loss) ### Tree: Enter position ├─ Rise (60%) → +$100 └─ Fall (40%) → -$50 EV = (0.6 × 100) + (0.4 × -50) = 60 - 20 = +$40 Wait └─ No position → $0 EV = $0 Conclusion: Entering position has positive EV (+$40), better than waiting ($0). ### Method Limitations ⚠️ Critical points: Subjective estimates — probabilities often "finger in the air" Doesn't account for risk appetite — ignores psychology (loss aversion) Simplified model — reality is more complex Unstable — small data changes can drastically alter the tree May be inaccurate — other methods exist that are more precise (random forests) But: The method is valuable for structuring thinking, even if numbers are approximate. ### 1. Structuring Ask: What are the action options? What are possible outcomes? What are values/utility for each outcome? How do we measure value? (money, utility units, happiness points) ### 2. Probability Estimation Help estimate through: Historical data (if available) Comparable situations Expert judgment (user experience) Subjective assessment (if no data) ### 3. Visualization Draw tree in markdown: Decision ├─ Option A │ ├─ Outcome A1 (X%) → Value Y │ └─ Outcome A2 (Z%) → Value W └─ Option B └─ Outcome B1 (100%) → Value V ### 4. EV Calculation For each option: EV_A = (X% × Y) + (Z% × W) EV_B = V ### 5. Recommendation Option with highest EV = best choice (rationally). But add context: Risk tolerance (can user handle worst case) Time horizon (when is result needed) Other factors (reputational risk, emotions, ethics) ### Trading & Investing Position Sizing: Options: 5%, 10%, 20% of capital Outcomes: Profit/loss with different probabilities Value: Absolute profit in $ Entry Timing: Options: Enter now, wait for -5%, wait for -10% Outcomes: Price goes up/down Value: Opportunity cost vs better entry price ### Business Strategy Product Launch: Options: Launch / don't launch Outcomes: Success / failure Value: Revenue, market share, costs Hiring Decision: Options: Hire candidate A / candidate B / don't hire Outcomes: Successful onboarding / quit after X months Value: Productivity, costs, opportunity cost ### Personal Decisions Career Change: Options: Stay / change job / start business Outcomes: Success / failure in new role Value: Salary, satisfaction, growth, risk Real Estate: Options: Buy house A / house B / continue renting Outcomes: Price increase / decrease / personal situation changes Value: Net worth, monthly costs, quality of life ### Operations Capacity Planning: Options: Expand production / outsource / status quo Outcomes: Demand increases / decreases Value: Profit, utilization, fixed costs Vendor Selection: Options: Vendor A / Vendor B / in-house Outcomes: Quality, reliability, failures Value: Total cost of ownership ### Calculator Script Use scripts/decision_tree.py for automated EV calculations: python3 scripts/decision_tree.py --interactive Or via JSON: python3 scripts/decision_tree.py --json tree.json JSON format: { "decision": "Launch product?", "options": [ { "name": "Launch", "outcomes": [ {"name": "Success", "probability": 0.4, "value": 500000}, {"name": "Failure", "probability": 0.6, "value": -200000} ] }, { "name": "Don't launch", "outcomes": [ {"name": "Status quo", "probability": 1.0, "value": 0} ] } ] } Output: 📊 Decision Tree Analysis Decision: Launch product? Option 1: Launch └─ EV = $80,000.00 ├─ Success (40.0%) → +$500,000.00 └─ Failure (60.0%) → -$200,000.00 Option 2: Don't launch └─ EV = $0.00 └─ Status quo (100.0%) → $0.00 ✅ Recommendation: Launch (EV: $80,000.00) ### Final Checklist Before giving recommendation, ensure: ✅ All options covered ✅ Probabilities sum to 100% for each branch ✅ Values are realistic (not fantasies) ✅ Worst case scenario is clear to user ✅ Risk/reward ratio is explicit ✅ Method limitations mentioned ✅ Qualitative context added (not just EV) ### Method Advantages ✅ Simple — people understand trees intuitively ✅ Visual — clear structure ✅ Works with little data — can use expert estimates ✅ White box — transparent logic ✅ Worst/best case — extreme scenarios visible ✅ Multiple decision-makers — can account for different interests ### Method Disadvantages ❌ Unstable — small data changes → large tree changes ❌ Inaccurate — often more precise methods exist ❌ Subjective — probability estimates "from the head" ❌ Complex — becomes unwieldy with many outcomes ❌ Doesn't account for risk preference — assumes risk neutrality ### Important The method is valuable for structuring thinking, but numbers are often taken from thin air. What matters more is the process — forcing yourself to think through all branches and explicitly evaluate consequences. Don't sell the decision as "scientifically proven" — it's just a framework for conscious choice. ### Further Reading Decision trees in operations research Influence diagrams (more compact for complex decisions) Utility functions (accounting for risk aversion) Monte Carlo simulation (for greater accuracy) Real options analysis (for strategic decisions) ## Trust - Source: tencent - Verification: Indexed source record - Publisher: evgyur - Version: 1.0.1 ## Source health - Status: healthy - Item download looks usable. - Yavira can redirect you to the upstream package for this item. - Health scope: item - Reason: direct_download_ok - Checked at: 2026-05-01T01:12:54.738Z - Expires at: 2026-05-08T01:12:54.738Z - Recommended action: Download for OpenClaw ## Links - [Detail page](https://openagent3.xyz/skills/decision-trees) - [Send to Agent page](https://openagent3.xyz/skills/decision-trees/agent) - [JSON manifest](https://openagent3.xyz/skills/decision-trees/agent.json) - [Markdown brief](https://openagent3.xyz/skills/decision-trees/agent.md) - [Download page](https://openagent3.xyz/downloads/decision-trees)