NW3:
I'll get a brief overview of my understanding of departures, written with a mostly FAA TERPS view, but also recognizing PANS-OPS. This is one of the most complicated and difficult areas to cover.
IF there is an IFR approach to an airport, it is supposed to have a surveyed procedure to leave IFR with obstacle clearance. I say supposed because I am sure there are exceptions for the usual aviation reason--money. If there is a published SID, follow it and abide by the published required climb gradient. On the chart or at the bottom of the airport diagram Jepp chart, in either feet/nm or percent gradient. You still have to clear the departure end of runway at the screen height, climb to 400 feet and then follow the SID with the required gradient. This is normal ops, does not cover the OEI case, which is another can of worms. There are more competent people here to answer that one.
If there is NO published SID, it is an "omni-directional" departure (PANS-OPS term, random departure TERPS) which requires a 3.3% (i.e.200 feet/nm) climb until reaching the minimum IFR altitude. Just clear the departure end of runway at 5 meters (PANS-OPS, 35 feet TERPS), climb straight ahead to 400 feet and turn on course.
The FAA, in the absence of an ATC requirement for a SID, will generate an obstacle departure procedure, if the rocks require something more than 200 feet/nm. It could be simple, "climb to xxxx MSL before turning on course" or a textual SID with radials, climb in holding, etc. etc. I am not sure if PANS-OPS does anything like a ODP, just a SID.
All that said, and it is about 80 pages reduced to this, nothing beats having a VFR chart, an idea of where the rocks are and EGPWS--all on them for me. Passing it off to the controller is OK, if he agrees to take the same punishment you receive for flying into a mountain.
GF