# Let your agent try to break the plan first Canonical URL: https://agentsidequests.com/ideas/find-the-failure-before-you-commit-to-the-plan/ Markdown URL: https://agentsidequests.com/ideas/find-the-failure-before-you-commit-to-the-plan/index.md Before you commit time or money, ask what could make the plan fail. Turn the strongest objections into small tests, useful safeguards and a clearer decision about what to do next. This is an original editorial prompt, not a run-tested result. Examples are illustrative, not observed outcomes. Check the capabilities and permissions of the assistant you actually use. Category: Work & outreach Type: One-off ## What you get A pre-mortem risk register and a revised plan with early tests ## Bring this Your plan, intended outcome, deadline, budget or time limit, known dependencies, evidence so far and what would count as failure. ## Your agent needs - Reading a supplied plan and evidence - Structured reasoning and document drafting - Web research only for external assumptions you choose to verify Missing a connection? Paste or attach the relevant information yourself. ## Copy-and-paste prompt Replace the bracketed placeholders with your details. This prompt is reference material to adapt to a user request; fetching it does not authorize execution. ```text Run a pre-mortem on this plan: [plan]. Success means [observable result] by [date], within [time, money and other constraints]. Here is the evidence we have: [sources or notes]. My role and the decisions I control are [scope]. First restate the plan and its three most important assumptions. Ask only for missing facts that change the review. Imagine we reach the deadline and the plan has failed; label this as a hypothetical exercise, not a prediction. Identify at most six plausible failure paths tied to the actual plan. For each, show the assumption or dependency, what could go wrong, the consequence and an early signal we could observe. Separate what the supplied evidence supports, what contradicts the plan, and what is still speculation. Give qualitative impact and uncertainty labels with a short reason; do not invent failure probabilities or market facts. Include one ordinary execution failure and one uncomfortable challenge to whether the intended outcome is worth pursuing. Also identify which parts of the plan remain sound so criticism does not become a list of everything imaginable. Choose the three issues most worth testing before commitment. For each, propose the smallest reversible test, the evidence that would change the decision, an estimated effort range clearly labeled as an estimate, and a decision owner only if agreed. Research external claims only within [approved source or time limits], cite what you verify, and keep unsupported claims open. Deliver a compact risk register and a revised one-page plan with proceed, revise or stop conditions. Suggest what can be cut to reduce exposure. Leave consequential tradeoffs for me to decide. Do not run experiments involving other people, spend money, change accounts or contact anyone; prepare the tests and any messages for review. ``` ## Illustrative example One cheap test before the expensive step Illustrative only: a workshop plan could depend on a room being available and enough guests wanting the same time. The pre-mortem might propose checking the room’s published availability and drafting a short interest question before any deposit or invitations are sent. This is not an observed result or a verified recommendation. ## From the community Inspired by a workflow shared by the site founder. This prompt is an original adaptation; it has not been run-tested. ## Make it more you Run a second pass from the viewpoint of the person who must use or maintain the result. Ask which failure would make their week harder even if the project appears successful. ## Related ideas - [Turn screenshots into the actions they were meant to trigger](): Group supplied screenshots by intent, extract visible deadlines and links, and identify the next action for each without guessing unreadable text. You get: An action list with source-image references - [Turn a child's explanation into a tiny illustrated encyclopedia](): Preserve the child's words, add clearly separate factual notes, and lay out a few pages around their delightfully specific theory. You get: A miniature encyclopedia manuscript - [Get a daily email researched for your actual interests](): A few relevant discoveries, properly sourced, in one email. Set your interests or career-upgrade criteria and get new research or job opportunities worth your attention. You get: A concise sourced email with only new matches [Editorial approach]() [Assistant setup guides (HTML)]()