I’m currently shipping a heavy feature that makes my product credit-based, depending on the tokens users actually use. Quite complicated and a lot calculations involved. Hmmm.. I should give it a try.
Oh, that sounds exciting and challenging. You might need something slightly more specific to your use case, but I think the skill here is a very good foundation. Give it a try.
Yeah, I’m still working on this feature, but I’m getting close to the end of development. Since it’s becoming pretty complicated, I really want to do some pressure testing like you mentioned in the article. Let’s see how it goes. I’ll definitely let you know if I run into any questions.
By the way, I learned something new while working on this: integer scaled storage for handling credits in the database. 🤠
nice! :) do you have maybe something similar for when the input is a screenshot of a chart (e.g. from a BI tool) together with the conclusions? I can imagine that might be useful, too.
quick question regarding this sentence in the skill: “Check for project context: read the CLAUDE.md”.
does it improve the outputs of the skill? I’m asking because http://CLAUDE.md is automatically included in the context at the start of a session, so I’d think Claude should already know this.
I’m currently shipping a heavy feature that makes my product credit-based, depending on the tokens users actually use. Quite complicated and a lot calculations involved. Hmmm.. I should give it a try.
Oh, that sounds exciting and challenging. You might need something slightly more specific to your use case, but I think the skill here is a very good foundation. Give it a try.
Yeah, I’m still working on this feature, but I’m getting close to the end of development. Since it’s becoming pretty complicated, I really want to do some pressure testing like you mentioned in the article. Let’s see how it goes. I’ll definitely let you know if I run into any questions.
By the way, I learned something new while working on this: integer scaled storage for handling credits in the database. 🤠
nice! :) do you have maybe something similar for when the input is a screenshot of a chart (e.g. from a BI tool) together with the conclusions? I can imagine that might be useful, too.
quick question regarding this sentence in the skill: “Check for project context: read the CLAUDE.md”.
does it improve the outputs of the skill? I’m asking because http://CLAUDE.md is automatically included in the context at the start of a session, so I’d think Claude should already know this.