What the workers can get away with will depend on quality assurance. No matter how they jack up their work, if the factory inspectors test the firearm and it doesn't work to spec, they'll scrap the bad parts and have them replaced. Sabotage prior to final inspection will only slow production. For the sabotage to have an affect out in the field, the sabotage has to be something subtle that doesn't show up with just a few test shots, something like poorly tempered springs.
First, let's look at this from the story telling angle. Why is the sabotage important to the story? What impact will it have? Is it necessary? Will the protagonist be saved because a minion has a sabotaged weapon? Will the sabotage be foreshadowed will complaints by soldiers, NCOs or officers? Or by showing workers sabotaging weapons at the factory? When the foreshadowing takes place determines how much time and detail is needed to make it work and believe me, you have to have foreshadowing. You can't have a soldier fire his weapon, miss and curse because the front sight was bent at the factory. You'll have your audience going "Where did THAT come from?"
If the front sight is bent, why wasn't it caught by the soldier it was issued to? Lack of training? Poor morale? Didn't the soldier shoot the weapon before this? If not, why not?
Unless you have a very specific scenario in mind, sabotaging the springs so they have a shortened life would work best. It's hard to detect during function testing, but chances the spring will fail increase with every shot and it can be frustrating to troubleshoot. For example, when an AR starts malfunctioning due to a failing extractor spring, the most common thought is that the AR is overgassed, or "Did you try a different magazine?"
I've always heard the stories about Nazi slave labor working to sabotage projects, but I've never seen anything that examined the impact it had on the Nazi war machine beyond slowing development of atomic weapons to a crawl. It would be a good idea to research what the actual affect was