Learning & Research
Analyze a Long Document With Citations
Answers your specific questions about a long document with quoted evidence and section references, and says what the document does not cover.
- Customizable fields
- 5 fields
Learning & Research
Builds a weighted decision matrix from your options and criteria, shows how sensitive the result is to your weights and flags missing facts.
Designed for general-purpose AI assistantsPublished November 2026
Fill in the fields and the prompt updates. What you type stays in this page and is never saved or sent.
Help me choose between options using a weighted decision matrix. I make the judgments; you make the reasoning visible.
The decision: {{decision}}
Options: {{options}}
What matters to me, and roughly how much (list in order of importance): {{criteria}}
Hard requirements (an option that fails one is out): none
Facts I have about each option (prices, specs, experiences, quotes): {{facts about the options}}
Do this:
1. Restate the decision, and remove any option that fails a hard requirement, explaining why.
2. Turn my criteria into a table. Give each criterion a weight from 1 to 5 based on the order and wording I used, and show me the weights so I can change them.
3. Score each remaining option on each criterion from 1 to 5, and give a one-line reason for every score. Base each score only on the facts I gave. If I gave no fact for a cell, do not guess: write "?" and list it as missing information.
4. Calculate the weighted totals and show the arithmetic. Count a "?" as unscored, and give totals both with those cells at 1 and at 5, so I can see how much the missing facts could change the result.
5. Test sensitivity: which single weight change would make the ranking flip, and does the leader stay the leader if my top criterion is weighted equally with the others?
6. Identify what the matrix cannot capture, such as risk, reversibility, relationships or how I will feel about the choice.
7. List the three facts that would most reduce my uncertainty, and how I could find each out cheaply.
8. Give a provisional recommendation, clearly labeled provisional and conditional on the missing facts.
Do not invent figures, reviews or specifications. Make the final choice feel mine: say what I would have to believe for each option to be the right one.4 required fields are empty and will stay as a {{placeholder}} in the copied text.
Illustrative only. Real answers depend on your input and the assistant you use.
Decision: choose a project tool for a 6-person team. Options: Tool A, Tool B, Tool C. Removed: Tool C fails the hard requirement "works offline". | Criterion | Weight | Tool A | Tool B | |---|---|---|---| | Easy to learn | 5 | 4 (your trial: set up in 20 min) | 3 (you said it needed a tutorial) | | Price | 4 | 3 ($8 per user) | 4 ($6 per user) | | Integrations | 3 | ? | 4 (lists your calendar) | Totals: Tool A 5x4 + 4x3 = 32, plus integrations unscored (+3 if 1, +15 if 5): 35 to 47. Tool B 5x3 + 4x4 + 3x4 = 43. Sensitivity: on the two criteria both tools were scored on, Tool A is 1 point ahead (32 to 31), so Tool B's lead depends on the unknown integrations score. Tool A overtakes Tool B (43) only if its integrations score is 4 or more (44 or 47). If Easy to learn were weighted the same as Price (4), the tools would tie on those two criteria (28 each), and Tool A would need an integrations score of 5 to lead (43 to 40). Not captured: switching cost from your current tool. Most useful facts: Tool A's integrations list, both tools' export options, and a one-week trial with two teammates. Provisional recommendation: Tool B, unless Tool A's integrations cover your calendar.
The matrix is a way to think, not an oracle. If the result surprises you, notice whether you are disappointed. That reaction is information about what you really care about, and a missing criterion.
Use it for decisions where the options are comparable. For a choice between very different kinds of things, the scores add false precision.
Save the final matrix in your notes with the date. When circumstances change, you can update the facts and see what moves.
Learning & Research
Answers your specific questions about a long document with quoted evidence and section references, and says what the document does not cover.
Coding & Software
Turns a technical decision and the options you weighed into an ADR with context, consequences and the conditions for revisiting it.
Learning & Research
Builds a realistic study plan from your deadline, available hours and current level, with checkpoints and a fallback for missed days.