2.5k
GitHub Stars
2
Bundled Files
2 months ago
Catalog Refreshed
3 months ago
First Indexed
Readme & install
Copy the install command, review bundled files from the catalogue, and read any extended description pulled from the listing source.
Installation
Preview and clipboard use veilstrat where the catalogue uses aiagentskills.
npx veilstrat add skill openclaw/skills --skill zeelin-us-iran-forecast- _meta.json305 B
- SKILL.md13.7 KB
Overview
This skill collects, verifies, and forecasts war and geopolitical conflict developments using authoritative open sources, cross-validation, and scenario-based prediction. It enforces strict separation of confirmed facts, assessments, and forecasts, and emphasizes explicit uncertainty, time-boxed forecasts, and trigger/disconfirming signals. The goal is disciplined, evidence-led intelligence synthesis for open-source analysis, not classified reporting or propaganda.
How this skill works
The skill builds a fact-first timeline from Tier 1 sources (Reuters, AP, official primary releases) and supplements with hard-data indicators (satellite reporting, shipping traffic, market moves). It tags each claim with a verification label (Confirmed, Single-party claim, High-confidence inference, Unverified) and separates battlefield, capability, political, regional spillover, and forward-indicator buckets. Forecasts are produced for three horizons (24 hours, 3–7 days, 2–6 weeks) with supporting evidence, constraints, confirming signals, and disconfirming signals.
When to use it
- Rapid synthesis after a new incident or escalation to establish verified facts
- Regular situation updates during an ongoing conflict (daily or multi-day cadence)
- Decision support when policymakers need probability-weighted near-term forecasts
- Monitoring spillover risks to markets, shipping, or neighboring states
- Preparation of scenario-based briefings for humanitarian or operational planners
Best practices
- Always separate facts, assessment, and forecast in outputs
- Build the backbone with Reuters/AP/official statements before using social media leads
- Use concrete dates and explicit time windows for all events and forecasts
- Label major claims with the mandatory verification categories
- For each forecast list supporting evidence, constraints, confirming and disconfirming signals
- Keep language probabilistic (high/medium/low probability) and avoid theatrical certainty
Example use cases
- Produce a dated 10-event timeline after a cross-border strike with source confidence labels
- Forecast likely retaliatory patterns over next 24 hours and list immediate watchlist indicators
- Assess whether a campaign is shifting to underground/strategic targeting and what constrains it
- Estimate regional spillover risk to shipping and oil markets over the next 2–6 weeks
- Create scenario-tree end-state probabilities with triggers that would move each branch
FAQ
It prioritizes Reuters, AP, and official primary releases first, then hard-data indicators, major international outlets, and uses social media only as early leads requiring verification.
How are forecasts expressed?
Forecasts are time-boxed into next 24 hours, next 3–7 days, and next 2–6 weeks and expressed in probability bands with explicit confirming and disconfirming signals.