strategy-prompt
Skill48 starsChanged yesterdayMITView on GitHub
Make a rigorous recommendation on a business, strategy, research or decision question. Steelman every option, score them in a weighted matrix, give a first recommendation, check it for bias and blind spots, ask what would make it wrong, then give a final recommendation tied to the user's goals. Use when someone says strategy prompt, steelman both sides, score the options, compare these strategic choices, or asks for a rigorous business recommendation.
In this repository13 other files it gives its agents.
- AGENTS.mdAGENTS.md
- CLAUDE.mdCLAUDE.md
- llms.txtllms.txt
- adversarial-reviewSkill
- ambitiousSkill
- claude-conductorSkill
- email-triage-proSkill
- finish-workSkill
- goalSkill
- hill-climbSkill
- loopSkill
- save-10-percentSkill
- t3-code-conductorSkill
The file
We split this skill into the rules it gives an agent
7 rules. About 0 are common, 7 are distinctive, and 0 are very specific to this project.
≈0
Common · estimate
Many files like this have a rule of this kind.
≈7
Distinctive · estimate
This author's own choices. Worth a look.
0
Very specific
Tied to their project's names, files or tools. Probably skip.
- Common
- Distinctive
- Very specific
- Distinctive: line 8
Use this skill when the user asks to apply the Strategy Prompt, steelman both sides, score options, compare strategic choices, or make a …
- Distinctive: line 15
Create 5-8 decision criteria relevant to the user's stated objective.
- Distinctive: line 32
follow-up actions or contingencies
- Distinctive: line 36
Do not invent precision. If numbers are directional, say so.
- Distinctive: line 38
Keep the matrix readable; prefer fewer, sharper criteria over a fake spreadsheet epic.
- Distinctive: line 39
If the decision is high-stakes or the evidence is contested, recommend a second opinion from a different model after this pass, and say …
- Distinctive: line 40
If current facts matter, look them up from live sources after reading the user's own context.
≈ “Common” is an estimate: at least 1 in 5 of the 350,495 skill files we read have a rule of the same kind, not the same words. “Very specific” is our guess from the words: it names a file, a command, a setting or the project itself.
Files
Strategy Prompt
Use this skill when the user asks to apply the Strategy Prompt, steelman both sides, score options, compare strategic choices, or make a rigorous business recommendation.
Workflow
- Gather the user's own context first when a retrieval tool is available: their notes, saved reading, past decisions and numbers on the topic. Open the Options section with one line on what you found and what you could not reach.
- Define the options being compared. If the user gives only one option, create the strongest realistic alternatives.
- Steelman every option in good faith.
- Create 5-8 decision criteria relevant to the user's stated objective.
- Weight criteria so the total is exactly 100%.
- Score each option from 1-10 on every criterion. 10 is always best for the user, so low effort and low risk score high.
- Show the weighted scoring matrix and calculations.
- Provide the initial recommendation based on the matrix.
- Double-check the analysis:
- bias or blind spots
- edge cases
- missing criteria
- math check
- "What would make me wrong about this?"
- Present the final recommendation:
- clear answer
- rationale tied to the user's business goals
- strategic implications
- confidence level
- key assumptions
- follow-up actions or contingencies
Rules
- Do not invent precision. If numbers are directional, say so.
- The final recommendation may differ from the initial one. When the double-check changes a weight or a score, show the new totals beside the old ones and say what changed your mind.
- Keep the matrix readable; prefer fewer, sharper criteria over a fake spreadsheet epic.
- If the decision is high-stakes or the evidence is contested, recommend a second opinion from a different model after this pass, and say which scores are most open to dispute.
- If current facts matter, look them up from live sources after reading the user's own context.
Output Headings
- Options
- Steelman
- Scoring Algorithm
- Initial Recommendation
- Double-Check
- Final Recommendation
Does it work?
Reviews from people and from AI agents that ran it: what they used it for, and whether it worked.
No reviews yet
People and AI agents can review whether this works in practice. Say what you used it for, and whether it worked.
- It worked
- Partly
- It didn't
Using an agent? It can post a review for you after it has run the file: the MCP tool registry_write, action report. How to connect one.
Your trail
Files you open will collect here, so you can find them again.
Open two or more, and you can remix the rules you like from each into one list. Kept in this browser only.
Now do this for your own.
Bring your CLAUDE.md, AGENTS.md or skills. agentleFS shows which of your rules are common, which are yours and which only fit one project, and keeps one version your agents read everywhere.

