This is just a guess. If you use an excess of salt, you are producing mild aqua regia and the breakdown products are yellow and, yellow + blue = green. I have noticed that when I use HCl to drop AgCl from copper bearing nitric solutions, it always turns yellow at the spot where I add the HCl. I have also noticed that, if I add a great excess of HCl to the nitric, the solution turns greenish.
I would dissolve the salt before adding it and add (in increments) just enough to precipitate all of the silver as silver chloride. I would stir it briefly after the additions. You can test for completion by letting it settle a bit from the surface and adding a single drop of salt water to the surface of the solution. If it produces a white cloud where you dropped it, you need more salt water. To get an idea of how much salt it will take, 1 gram of pure dry salt, when dissolved, will precipitate 1.84 grams of silver.
I prefer using HCl instead of salt. Both rock salt and table salt contain solid impurities which would end up being mixed with the AgCl and this may, or may not, be undesirable. One ml of 31% muriatic acid will precipitate about 1.05 grams of silver. If you're using 37% HCl, 1 ml will precipitate about 1.29 grams of silver.
Also, if you know, fairly accurately, how much HCl you have added, without using an excess, you can easily make a calculation from the above figures and get a good idea of the amount of silver you have precipitated. The same way with the salt. I have also used this method to roughly titrate the silver in a sample of a silver nitrate solution, in order to get an idea of how much silver was in there. On a more precise scale, there is a old, standard, very accurate, method of silver analysis called the Guy Lussac method that is based on essentially the same thing.
