Install
openclaw skills install @wlkqyang-star/zeelin-us-iran-forecastCollect, verify, and forecast war or geopolitical conflict developments using authoritative sources, cross-validation, and scenario-based prediction.
openclaw skills install @wlkqyang-star/zeelin-us-iran-forecastTurn a request like:
into a stable workflow that is:
This skill is for open-source intelligence style synthesis, not classified intelligence, not propaganda, and not fantasy war writing.
Every answer must be split into:
Never present a forecast as if it were already confirmed reality.
In war, every side has incentives to:
Therefore:
Do not write:
Write:
Do not sound certain when evidence is thin. Prefer:
For each prediction, state:
For any armed conflict or fast-moving geopolitical crisis, collect at least these five buckets:
Build the answer from the top of this hierarchy downward.
Use these first to build the main factual structure.
Use for:
Use for:
Use to confirm what an actor said, not automatically whether the claim is true:
Use where available:
Use to add depth, background, or insider context.
Use for:
Rule:
Use for:
Rule:
Use only for:
Never treat them as settled fact until checked.
If using them, verify:
Use for:
Do not let commentary substitute for hard confirmation.
For every major claim, implicitly or explicitly classify it as one of these:
Confirmed
at least two reliable sources or one reliable source plus strong corroborating data
Single-party claim
one side says it; independent confirmation is missing
High-confidence inference
not formally confirmed, but several indicators point the same way
Unverified / preliminary
not safe to treat as established
Examples:
Before details, answer:
Build this first with Reuters + AP + official primary sources.
List at least 5–10 major events with:
Format:
Always separate what each actor wants.
Do not merge these into one “allied side” goal.
Prediction quality depends on knowing not just intent, but limits.
Check:
Never give one undifferentiated prediction blob.
Look for:
Look for:
Look for:
For each forecast, use this structure:
Example:
List the supporting signals:
List what could slow or limit this:
Do not write one cinematic ending. Use scenario trees.
Typical signs:
Typical signs:
Typical signs:
For each scenario:
Use this order:
Example:
Group into:
Explain:
Split by:
Within each, separate:
At least three:
Give 3–7 indicators that would change the assessment fastest.
Fix:
Fix:
Fix:
Fix:
Fix:
Before delivering, verify:
Be slower than propaganda, but more accurate than hype.