recipe
Validate LLM output before it propagates
· 2 min read
A bad structured response exits 1 before it reaches downstream systems, so you catch schema drift at the source. Check the LLM response against a ...
tag
#prompt
11 posts tagged #prompt
recipe
Token-checked LLM call
· 2 min read
If the prompt is over the budget, the pipeline stops and the model never sees a prompt that only half fits. Count the tokens in a prompt and stop the ...
recipe
Scrub secrets from LLM output
· 2 min read
A model can echo a secret back, and mask removes it before it reaches storage. Redact secrets from an LLM response before the pipeline stores it in ...
recipe
Retry flaky API call
· 2 min read
A transient 500 does not stop the run, because coax retries the call with backoff. Retry an LLM call with exponential backoff, and fail the attempt ...
recipe
Preview LLM request
· 2 min read
You see the exact request, after masking, before any money or secrets leave your machine. Add --explain to vrk prompt to print the request as a curl ...
recipe
Fetch and summarise a page
· 1 min read
If the page is too long, the pipeline stops before the API call, so you do not pay for a request that cannot fit. Fetch a web page as markdown, check ...
recipe
Diff review budget
· 2 min read
A diff that is too big fails fast and cheap, instead of being cut off halfway through the review. Check the token count of a git diff, then send it ...
recipe
Daily token budget
· 2 min read
The counter refuses the call once the daily limit is reached, so a runaway loop cannot use up your API budget. Add each prompt's token count to a kv ...
recipe
Cache LLM response
· 2 min read
The content hash is the cache key, so a rerun with the same input makes no API call. Hash the input, look the hash up in kv, and call the model only ...
recipe
Batch LLM with rate limiting
· 2 min read
throttle releases filenames at a fixed rate, so each API call stays inside the rate limit. Process a large set of documents at a fixed request rate, ...
recipe
Agent-safe pipeline with secret masking
· 2 min read
If an agent handles user data, mask it first, so secrets never reach the LLM or the kv store. Mask secrets, check the token budget, call the LLM, and ...