This occurred to me as a quick-and-dirty way to put some of those non-adventuring skills back into the game, assuming they haven't already done so and not told us. I'm making certain assumptions about the way rituals work, but they seem reasonable given what we know.
Craft takes time and money (raw materials) and produces a result. Rituals take time and money (components) and produce a result. Why not model the former with the later?
Weaponsmithing
cost: half weapon cost
time: 1 week (or varying time with weapon, whatever you like)
requirements: forge
result: produces a mundane weapon
The DM can grant access to this in a variety of ways, depending on his campaign preferences. This could include taking "Crafting" as a feat, or it could simply be granted as a result of game play or downtime practice. Gaining further craft "rituals" could cost time and training rather than gold. The crafting feat might grant access to the magic item-making rituals as well, if your campaign includes craftsman-made magic items rather than only caster-made items.
Since these things are rituals, rather than skills, you don't run into the issue of trading adventuring effectiveness for non-adventuring things. The issue of how broad to make a field of crafting is also not a big deal (wait, I need ironwork and leatherwork to make studden leather?) since you can freely learn more crafts with just time/gold/effort.
I can imagine other kinds of non-magic rituals as well:
Dress to Kill
cost: X gp (being fashionable is expensive)
time: 1 hour
result: grants +2 on diplomacy and bluff checks for the next day
Naturally the king has people to handle this for him every day, so your +2 and his +2 cancel out (assuming diplomacy vs diplomacy) when you get your audience with him. So maybe you need the higher level (and more expensive) version that gives +4. And if you don't have it, you're at -2 effectively. This could be a simple way to model social status without inventing a new stat to track.
Craft takes time and money (raw materials) and produces a result. Rituals take time and money (components) and produce a result. Why not model the former with the later?
Weaponsmithing
cost: half weapon cost
time: 1 week (or varying time with weapon, whatever you like)
requirements: forge
result: produces a mundane weapon
The DM can grant access to this in a variety of ways, depending on his campaign preferences. This could include taking "Crafting" as a feat, or it could simply be granted as a result of game play or downtime practice. Gaining further craft "rituals" could cost time and training rather than gold. The crafting feat might grant access to the magic item-making rituals as well, if your campaign includes craftsman-made magic items rather than only caster-made items.
Since these things are rituals, rather than skills, you don't run into the issue of trading adventuring effectiveness for non-adventuring things. The issue of how broad to make a field of crafting is also not a big deal (wait, I need ironwork and leatherwork to make studden leather?) since you can freely learn more crafts with just time/gold/effort.
I can imagine other kinds of non-magic rituals as well:
Dress to Kill
cost: X gp (being fashionable is expensive)
time: 1 hour
result: grants +2 on diplomacy and bluff checks for the next day
Naturally the king has people to handle this for him every day, so your +2 and his +2 cancel out (assuming diplomacy vs diplomacy) when you get your audience with him. So maybe you need the higher level (and more expensive) version that gives +4. And if you don't have it, you're at -2 effectively. This could be a simple way to model social status without inventing a new stat to track.