Gulf + Middle East Hybrid Intelligence Skill
This file is the root entry point for agent discovery.
Two platform-specific variants are available — choose the one matching your environment:
| Environment | File |
|---|---|
| Claude | runtimes/claude/SKILL.md |
| Codex | runtimes/codex/SKILL.md |
Both variants share the same analytical contract: mechanism-first reasoning, evidence labels (Verified / Plausible / Judgment / Unknown), Iran actor distinctions (Iran-state / IRGC-affiliated / Iran-private commercial), role-based implications, trigger points, confidence footer, and explicit limitation notes. Platform additions are documented inside each file.
For project identity, scope, evidence rules, and agent behavior contract, read AGENTS.md.
For current status against the Definition-of-Done bars, read STATUS.md.
For the primary claim/source handoff to the Agenda Intelligence MD evidence-packet linter, read docs/evidence-packet-handoff.md.