BA what-if-scenario-builder
Structured methodology for constructing rigorous "what if?" scenarios combining foresight frameworks (Shell, Schwartz, Manoa, Intuitive Logics), deep analytical tools (CLA, Cross-Impact, Morphological, Counterfactual History, Futures Wheels, Analogy Search), stress-testing techniques (Pre-Mortem, Red Teaming, Base Rate Check), and quantitative patterns (Monte Carlo, Agent-Based, Backcasting). Includes Divergence Depth selection, Temporal Cascade framing, and Agency Uncertainty principles. Use whenever the user asks "what if?", wants to explore hypotheticals, build scenarios, analyze alternatives, stress-test plans, or imagine how things could be different. Also trigger on: scenario planning, contingency analysis, alternative futures, speculative reasoning, "imagine if", "suppose that", thought experiments, futurecasting, counterfactual — even without explicit "what if."
As a process A 86/100 · Runs to the end — weak spots: execution cost, progress reporting
How to improve
- The SKILL.md body is over 5,000 tokens: move reference detail into references/ and load it when needed.
- Your own cases (evals/evals.json, 4–6 real requests with expected answers): the full check would then run those instead of a model-drafted suite.
- A spec.yaml with trigger phrases and assertions — a behaviour contract for CI; `skilltest init` writes a template.
Guard findings · 5
✓ No critical or high findings
Medium and low: 5
-
low Risky intent
intent-offensive-securitySKILL.md:7Offensive-security / dual-use content (legitimate for authorised testing; review intended use)(Pre-Mortem, Red Teaming, Base Rate Check), and quantitative patterns (Monte Carlo, Agent-Based,
-
low Risky intent
intent-offensive-securitySKILL.md:806Offensive-security / dual-use content (legitimate for authorised testing; review intended use)#### 15. Red Teaming / Adversarial Thinking
-
low Risky intent
intent-offensive-securitySKILL.md:816Offensive-security / dual-use content (legitimate for authorised testing; review intended use)Red teaming is not cynicism — it is intellectual hygiene. A scenario that survives aggressive red
-
low Risky intent
intent-offensive-securitySKILL.md:823Offensive-security / dual-use content (legitimate for authorised testing; review intended use)Red Team challenges:
-
low Risky intent
intent-offensive-securitySKILL.md:839Offensive-security / dual-use content (legitimate for authorised testing; review intended use)Red Team conclusion: The scenario is plausible but depends on three things breaking right
Files scanned: 1. Evidence is masked. Grey chips explain why severity was lowered.
Against the Agent Skills spec
- warning
body-longSKILL.md body ≈ 19055 tokens (recommended < 5000); move details to references/
Process rating: all ten parameters 86/100
- 0Progress reporting. Says nothing while it works
- 10Execution cost. Instruction body is 19055 tokens: crowds the task out of the window
- 70When it triggers. States when to use, but not when not to
- 70Inputs and preconditions. Inputs and preconditions are listed
- 100Tools and files. No external tools needed
- 100Steps. 209 steps
- 100Result and completion. Output format and completion criterion are stated
- 100Failures and branches. 19 branches, has a failure section
- 100Consistency. Name and required fields are in place
- 100Running it twice. Mutating operations check current state
- medium Safety rules and hard prohibitions inside a skill: they belong in the system prompt, here they protect nothing
Everything here is measured from the skill text rather than judged by a model, so the numbers are checkable. A parameter weighs more when it is a more common reason for the process to stall.
Quality signals
- +4Description does not say when NOT to use the skill (false activations)
- +3Description length 882: 120–800 characters recommended
- -221 emoji in the instructions: noise for the model
- +1No license
- +2Single-language instructions
- +5Description quotes 5 example trigger phrases
- +4Structure: 39 headings
- +3Step-by-step instructions: 209 items
- +3Output format is stated explicitly
- +4Has examples (2 code blocks)
Quality base 70; lint remarks subtract, signals add up to 100. Result: 77.