Germany kept selling Israel arms. Just not for Gaza.

Berlin stopped approving military equipment that could be used in Gaza, while leaving the wider German-Israeli defence relationship intact. That distinction made the policy narrower than an embargo but more immediate than a diplomatic rebuke. Intended use became a condition of export.

Israel kept its existing systems; future consignments within the defined category faced a German decision. A supplier government does not need to close every route to make one operationally uncertain.

The obvious asset is what the buyer can photograph. The decisive dependency is often selective approvals, delayed paperwork, mission limits, temporary pauses and decisions applied to one product but not another. That is how policy is reduced to the simple labels of sale, embargo or cancellation. That is the control point hiding behind the headline fact: Germany kept selling Israel arms.

The most dangerous failures are often deniable. A component is unavailable, an approval takes longer than expected or another customer becomes urgent. Each explanation can be reasonable. Together they can create the operational effect of a refusal while avoiding the political cost of saying no outright. For Israel, the warning is captured by what followed: Just not for Gaza.

Not every pause is an embargo and not every delay is political. But the force still has to operate while lawyers and diplomats decide what the pause should be called. That distinction matters here: Just not for Gaza.

For Israel, a government does not need to say no when 'not yet' works just as well. That is what 'Germany kept selling Israel arms' should force a buyer to test.

Do not ask whether the supplier is friendly. Ask questions that can be evidenced. Are spares and munitions treated differently from complete systems? What remains permitted that could provide a practical workaround? How long can the force operate while the policy remains deliberately ambiguous? That is what 'Germany kept selling Israel arms' should force a buyer to test.

Run the failure as a tabletop exercise before it becomes a real one. Assume the controlling party decides to create the operational effect of a refusal while avoiding the political cost of saying no outright. Who notices first? Who has authority to activate the alternative? How many days of usable stock remain? If the answers require a new committee, a new budget and a new contract, you do not yet have a fallback. The exercise has a concrete starting point in this case: Germany kept selling Israel arms.

The buyer still has options: read the scope of the restriction rather than the political headline, separate complete systems, munitions, spares, training and support in every risk review and prepare alternatives for an indefinite pause, not only a formal embargo. These measures are rarely as exciting as the platform. They are often what make the platform usable. That is what 'Germany kept selling Israel arms' should force a buyer to test.

Then test the fallback. Exactly which licences, products and missions are affected? Is the restriction written, temporary or merely administrative? Who has authority to lift it, and on what evidence? A plan that exists only in a slide deck will not become real because the crisis has arrived. That is what 'Germany kept selling Israel arms' should force a buyer to test.

The final test for Israel is simple: what military task becomes impossible if an administrative pause lasts six months rather than six days? If the honest answer begins with 'provided that somebody else agrees', that condition belongs on the first page of the procurement decision, not in an annex nobody reads. That exposure sits behind the claim that 'Germany kept selling Israel arms'.

Ambiguity benefits the state applying the pressure. The buyer carries the uncertainty, the readiness loss and the cost of waiting. The most effective veto may never be announced as one. That exposure sits behind the claim that 'Germany kept selling Israel arms'.