Hello, had a look at your map and I think it has a lot of potential. I looked through the nav in a listen server and have made some modifications, primarily converting two-way connections to one-way dropdowns where appropriate (ie where bots can fall off a square in one direction, but would be unable to jump up the other way) and removing small unecessary squares on fences and crawl spaces, and would also like to suggest that you don't pack the nav into the map for future versions (as it will allow for the nav to be updated without a new map version; this is the standard for community MvM).
drive.google.com
(You will also have to run a nav_analyse yourself, as this is impossible to properly do on a map with packed nav.)
Overall, The nav as it is didn't have any particularly glaring issues, I wouldn't be surprised if you did some cleanup to it after generating it. I would recommend you run around your map with ent_fire !self setmodelscale 1.75 in order to find any spots which a sentry buster would be unable to access, and either prevent engineers from building in those areas or moving up the ceiling/widening the corridor. Also keep in mind that
the bomb can land anywhere with nav, regardless of whether robots can actually reach that area or not. If you modify some parts of the map to circumvent this you will have to re-nav them because I removed nav where bots cannot naturally move.
Some issues you will need to determine a fix yourself for are as follows:
^ Sentry Busters will struggle getting under this roof if they come from this side, raising the ceiling a bit should alleviate this.
^ Despite the impracticality of this as a sentry spot, sentry building should still be disabled in this shack, as busters are completely unable to enter it, which would cause them to detonate when they spawn if a buster spawns as a sentry is built here. AFAIK there's a way to filter func_nobuild to sentries only, as this would be a pretty neat spot for a teleporter.
^ Giants cannot enter this building from this side without crouching, which they are not naturally inclined to do, and will wait for several seconds before doing so. I recommend you increase the height of the gate. Remember to test it with ent_fire !self setmodelscale 1.75.
^ This area is quite unfortunately designed, with nav a bomb can get stuck up here and prevent the bots from winning; but without nav this could be quite a significant
godspot (a spot which bots have trouble seeing players, as bots can only detect players if they have a line of sight on the last nav square the player was standing on). You could either try to find a way to have nav there without the possibility of a bomb stuck; such as changing the layout or using some mapping magic I'm unaware of, or you could decide which issue you think is least pertinent (tip: it's the godspot which puts bots at less of a disadvantage).
^ Same issue.
^ Very unfortunate overpass, absolutely no idea how you're going to handle this one. The whole thing is too short for giants and is quite close to the action. Removing the nav isn't really an option and I'd assume it's a lot of work to raise the ceiling/lower the floor to allow giants access. You'll either have to live with it (and use a bot_giant tagged nav_avoid so giants don't attempt to cross under it).
My game crashed as I tried to check, so I will tell you just to be safe. You need to generate/analyse your nav with a red func_respawnroom in red's spawnroom, otherwise robot Spies will have a chance to spawn in there. (This can be seen on MvM Production).
As I said at the start. Your map has potential, only some minor kinks to work out. This was very clearly tested well. I hope you return to MvM mapping in future, this map has a rather unique style to it without straying awfully far from what makes good MvM maps good. Also the documentation missionmakers are given is amazing, actually visualising the spawns and paths instead of just saying they exist.
NOTE: If you wish to discuss with me further, I recommend you add me on Steam or Discord (randomguy3016) as I will be much faster to respond on there than on here.