A bunch is eight arranged in a circle 2" (5cm) in diameter. I've already run into some difficulty wiring these so I'd like to share my plans here because I've gotten some very helpful recommendations.
The LEDs are glued into holes in an acrylic box so the base portion is slightly above the surface of the acrylic. They are mounted on the underside of the lid so some movement between LEDs and connection to a PCB is expected, mostly during construction.
Plans for the supply lead are to bend the leg of the LED so it is parallel to the underside of the lid and a few mm above it. I will solder stranded wire to the leg and then cover the solder joint with some shrink wrap. I also plan to put a dab of hot melt glue on the base of the LED to act as a strain relief for the legs.Near the hinge I will snub the leads to the lid so there is no movement at the LED. (lead, lid, LED... did I actually type all of that correctly on the first try? ;D ) I will run the leads around the circle, binding them together either with more shrink wrap or zip ties (or perhaps some cable lacing.)
I could do the same with the ground leads, but it seems like a waste of wire since they are all ultimately connected together. I'm thinking about a decent way to provide a common ground connection to all eight LEDs. The first attempt included a bare wire bent in a circle and soldered to the ground leg of each LED. A connection to the one on the end and run to a ground on the processor completes this part of the circuit.
Is a bare ground wire OK? One alternative is to use insulated wire and strip a small section at each LED to expose some bare wire for soldering. That would reduce the exposure of the ground wire. I could paint the solder connection with some liquid electrical tape to further reduce exposure. (It might make sense to paint the supply side of the LED with that too.)
In addition, there are four push-buttons outboard of ring of LEDs which also require ground connections. (The other connection on each push-button goes to a digital input configured with the internal pullup resistor enabled.)
I'm interested in hearing about other strategies that have been employed for similar ground connections. Surely others have faced this same issue.