Switched to @zeddotdev two weeks ago after their 1.0. Was interested in speed and got it. Started off by disabling all AI features in editor and using it to tweak/review output from Claude. So far so good!
If I want to use @Cloudflare to reverse proxy a folder then I _must_ either a) use workers b) have an Enterprise account to set Origin Rules / Page Rule Host headers? Why is this so hard!? Am I missing a doc that explains this?
If you don't want to work against the constraints of the Anthropic APIs you must keep your system prompt and tools fixed across sessions. Push customer specific instructions down and design around not changing tools.
Useful tip to cut time-to-first-token on longer prompts in the API: pre-warm the prompt cache.
Send your system prompt before the user prompt. Claude writes it to the cache, but skips generating any output.
When the real user request lands, it'll hit a warm cache.
@tirellat@Shopify I did โ the specific answers don't exist in the docs and what is in the docs only includes a subset of the behaviour. Perhaps this is intentional in that the unspecified behaviour is considered internal / private but it would be good to say so!
I'm always curious about behaviour of HTTP proxy features in web apps. @Shopify's App Proxy forwards HTTP requests to app defined origin. Docs are light on constraints โ connect / read timeouts, response size, Shopify added headers (x-shopify) โ any authoritative sources on this?
@randomfrequency@claudeai Claude Code on the Web doesn't have `gh` (at least for us?) so I had a startup script that injected a PAT and told it sternly "YOU DONT HAVE `gh` SO U MUST USE THE API w/ THIS TOKEN". Glad to remove that. Would be nice if the docs mentioned it tho.
code mode: let the code do the talking
(aka, after w/i/m/p)
wherein I ponder the implications of every user having a little coding buddy, and every "app" being directly programmable on demand.
https://t.co/i62HuWsjKg
lmk what you think.
With Anthropic server tools web search / fetch, I find it necessary to include the # of allowed domains & calls in the prompt to avoid tool call errors. Is it just me running into this?