Your LLM Proxy Retries the Wrong Errors
What an LLM proxy has to decide on every failure: limits are per organization, which statuses to retry, why nothing is retried after the first streamed byte, and the headroom headers the API already sends.
#prompts-guide
12 posts
Difficulty
Your LLM Demo Works. What Every Call Needs in Production
One timestamp in the system prompt and every call misses the cache. Pricing each call, making failures loud, logging per feature, and treating outside text as untrusted.
RAG That Works: Why Retrieval Matters More Than the Prompt
Most bad RAG answers are retrieval problems. How RAG works end to end, the three parts of the grounded prompt, how to measure retrieval alone, and when long context is enough.
Did Your Prompt Change Help? Prove It with an Eval
A model upgrade can change behaviour with no error. What an eval is, the three graders, the held-out slice, and how evals make upgrades routine.
Build the Agent Loop Yourself, Once
Tool use and the agent loop in eleven lines of Python, the four changes that harden it, and when one loop stops being enough.
Your Prompt Worked in Chat. Why Is the LLM API Different?
A chat prompt moved into code goes flat, forgets and fails with 400s. Send the history, write the system prompt, set effort and a schema, and read the stop reason first.
Turn Repeated Fixes into Skills, Then into a Team Practice
A correction typed into every session should become something that runs without you. A skill for API tests with a small eval and five pillars, a one-page team policy measured against a baseline, and four levels to place yourself.
Prompt Injection and the AI Bill: Give the Agent Less to See
One line in an issue can make an agent send your secrets out, and a long session bills the same conversation many times. Break one leg of the lethal trifecta, make secrets unreadable, and clear context between tasks.
AI-Generated Code: Checks That Can Fail, and the Calls That Stay Yours
AI-generated code fails in ways that look fine, and people rate it higher than it deserves. How to prove a test can fail by breaking the code, what to read in a diff, and which decisions stay with people.
Run Coding Agents Without Babysitting, and Reset Them When They Loop
A brief written like a ticket, a check the agent can run, a plan you approve before code, and fences set in configuration. When it goes wrong, name the pattern and change what it sees.
Rules Files and a Prompt Library: Make the Repo Remember
If you type the same thing in two chats, it belongs in the repo. What earns a line in a rules file, a filled-in AGENTS.md, and a shared prompt library whose review prompt stops hiding real bugs.
How LLMs Work, and How to Ask So They Stop Guessing
A model predicts the next token from its context window and nothing else. That explains confident wrong answers and the fixes: put the source in the window, give it an out, and write the prompt as a spec that answers five questions.