chr15m

12 ms·

271.
by chr15m 10mo ago
The true path to AGI.
272.
by chr15m 10mo ago
echo "rm -rf ~/ > safe-rm" chmod 755 safe-rm ./safe-rm Sandboxes are hard, because computer science.
273.
by chr15m 10mo ago
It behaved exactly like 99% of developers, introducing unnecessary complexity.
274.
by chr15m 10mo ago
Take my money! Literally take it, I just ordered. I have been imagining this exact device existing and now it does, yay, thank you!
275.
by chr15m 10mo ago
In the notes it says you can configure where it goes.
276.
by chr15m 10mo ago
Sharing it in your blog over a period of months or years is how you become a known figure eventually.
277.
by chr15m 10mo ago
Wish I could upvote this harder. Thank you arXiv!
278.
by chr15m 11mo ago
> deleted my production database I'm astonished how often I have read about agents doing this. Once should probably be enough.
279.
by chr15m 11mo ago
People blaming the user and defending the software: is there any other program where you would be ok with it erasing a whole drive without any confirmation?
280.
by chr15m 11mo ago
This is cool. An even more minimal bash version. Love it!
281.
by chr15m 11mo ago
I've now added caching to runprompt with the --cache flag and RUNPROMPT_CACHE env var. Thanks for the suggestion!
282.
by chr15m 11mo ago
I've implemented this now. You can set it with BASE_URL or OPENAI_BASE_URL which seems to vaguely be the standard. I also plan to use this with local LLMs. Thanks for the suggestion!
283.
by chr15m 11mo ago
llm stores data (prompts, responses, chats, fragments, aliases, attachment metadta) in a central sqlite database outside the working directory, and you have to use the tool to view and manipulate that data. I prefer a tool like this to defa
284.
by chr15m 11mo ago
I added this and you can now make .prompt files with a runprompt shebang. #/usr/bin/env runprompt
285.
by chr15m 11mo ago
Bash pipelines don't do any caching and will execute fresh each time, but I understand your idea and why a cache is useful when iterating on the command line. I'll implement it. Thanks!
286.
by chr15m 11mo ago
Will have a look at this. That could be the way to go. Thanks.
287.
by chr15m 11mo ago
I'm not sure. I think I need to use it more to see what the LLMs do with bad data. The design you're suggesting might be the answer though.
288.
by chr15m 11mo ago
Less typing. More control over chaining prompts together. Reproducibility. Running different prompts on different providers and models. Easy to install and runs everywhere. Inserts into scripting workflows simply. 12 factor env config.
289.
by chr15m 11mo ago
Cool, I'm going to add file modalities too. Thanks for the validation!
290.
by chr15m 11mo ago
I don't want to introduce a dependency. Simon's tool is great but I don't like the way it stores template state. I want my state in a file in my working folder.
291.
by chr15m 11mo ago
Great article, thanks. "One-liner code review from staged changes" - love this example.
292.
by chr15m 11mo ago
If the LLM returns invalid schema the next link in the chain will send that through since it defaults to string input if the JSON doesn't parse. Maybe I should make it error out in that situation.
293.
by chr15m 11mo ago
That's on my TODO list for tomorrow, thanks!
294.
by chr15m 11mo ago
Good idea. Will figure out a way to do this.
295.
by chr15m 11mo ago
Do you mean you want responses cached to e.g. a file based on the inputs?
296.
by chr15m 11mo ago
Oh interesting. Will investigate, thanks!
297.
Show HN: Runprompt – run .prompt files from the command line(github.com)
134 points by chr15m 11mo ago|49 comments
298.
by chr15m 11mo ago
lol
299.
by chr15m 11mo ago
Imagine using an all-in-one suite.
300.
by chr15m 11mo ago
With 250+ applications per job, does your company carry the prestige of Rolls Royce? Everything you are saying affirms the point OP is making.

More ›