I have to analyze the magic-user spell list in terms of the build-a-spell system I've been developing, to see if it's viable and to look for patterns. Might as well do it one level at a time, right here in the blog. I'm using Men & Magic, here.
Detect Magic can be explained easily. A detection cantrip would amplify or otherwise remove a restriction from an ordinary sense; Detect Magic adds a feature to this basic concept, letting the cast sense an otherwise undetectable magical aura. 1 Feature = 1st level spell.
Hold Portal could almost be a cantrip, duplicating the effects of spiking a door without needing to use a spike. Assume that it's 1st level because it's ranged: 1"/level or HD.
Read Magic depends on interpretation of what magic writing is: is it an incomprehensible set of symbols that can't normally be understood without magical assistance, or is it just a really complicated code that takes a long time decipher? The former interpretation works for a 1st level spell, but I actually prefer the latter, which makes this better as a cantrip: 1 turn of study with this spell is equivalent to 1 day of research without.
Read Languages is easy: 1 Feature (comprehend otherwise unknown language) = 1st level spell.
Protection from Evil is like armor with a few other features, but it only applies to enchanted and evil creatures. I was tempted to try to force an interpretation based on a hypothetical cantrip, but decided to just dump all variables and go with 1 Feature = 1st level spell.
Light builds on a hypothetical cantrip-level effect (candlelight without a candle) and adds a duration. 1 Variable = 1st level spell.
Charm Person guarantees an Extremely Friendly result without any reaction rolls for a minimum duration. 1 Variable = 1st level spell.
Sleep is the tricky one. If we assume a cantrip that accelerates the normal progression of drowsiness, we could either add a variable quantity (affect more than one target) or a variable range, or a variable duration, but not all three. Assume the first, force the MU to approach to 1" range, and let targets roll to wake up each round.