Nah, you're not breaking any rules.  It's actually a very good question.

Explosive compositions/compounds are classified as such because they produce a lot of energy upon decomposing, and most of their decomposition products are in a gaseous form which builds up a very high pressure as the reaction proceeds.  What determines the stability of something is the activation energy required to get the reaction going.  If it has a very high activation energy (like ammonium nitrate does), then you need to impart a good deal of energy to start getting it to decompose.  However, once it decomposes it will supply enough energy to keep going in a chain reaction leading to a detonation.  For something like nitrogen triiodide, however, the activation energy is very low.  All it takes is a little bit of energy in the form of a vibration or 'nudge' and the activation energy barrier is overcome resulting in a detonation.

What tends to lead to the low activation energy is weak intramolecular bonding, or highly stressed bonds.  In the case of NI3, the N-I bond is pretty weak and there is a LOT of strain on those bonds due to the incredibl large size of the iodine atom.  As a result, the atoms aren't being held together very tightly.  Meanwhile, the bonds created when the products form are VERY strong and are much more stable.  In nitroglycerine, you have three -O-NO2 groups on the glycerine backbone.  The C-O single bond isn't a very strong bond, and the NO2 groups are fairly large in size.  As a result the weak bonds and the sterics of the molecule make it very unstable.  Once again the products all have strong bonds so there's a huge difference in energy.  As a result, nitroglycerine can easily be influenced into decomposing, and once it starts going off the energy it releases sets off other molecules and so on and so on.  The chain reaction ensues and you get a huge explosion.

A basic way to determine if an explosive mixture will be stable or unstable is to look at the structure.  If there are multiple weak bonds and large atoms/groups attached at multiple points, it will probably be fairly unstable.  If there are very strong bonds and not too much steric hinderance, then the 'explosive' will tend to be more stable as the energy required to break a molecule down will be higher.

most of their decomposition products are in a gaseous form

Not necesarilly. IIRC hmx9123 some time ago gave heavy metal acetylides as a counterexample - they are strong explosives, yet there are only metal and soot in the reaction products.

Heavy Metal, and light metal, acetylides decompose into the metal and acetylene gas.

Where is the hydrogen source?

hmx post about acetylides explosion

Never realized that acetylide is another name for carbide.

I am not 100% sure about details, but there are different carbides - I have seen them referred to as methinides (?not sure about spelling) and acetylides, IIRC there are others too.