Stage 2 · Lesson 5 of 7
Tool errors the model can recover from
Tools fail: an order is not found, an API times out, an argument is invalid. Do not let the exception crash the loop. Return a tool result marked as an error, with Anthropic's is_error flag or a clear error field, and a short message that says what went wrong and what to try instead.
For example: 'No order found for that email. Ask the customer for the order number, which starts with ORD-.' A good error lets the model retry with fixed arguments, choose another tool or ask the user, which is most of what makes an feel robust. Log the full technical error for yourself separately.
Common trapreturning an empty result on failure, so the model concludes the customer has no orders and says so confidently.
A question to check yourself, an AI practice session and 1 cheat sheet entry come with the full roadmap. Unlock
Resources for this stage
Start with· follow one of these
- GuideFreeClaude Tool Use Documentation
Anthropic · Intermediate
Read Overview, Define tools and Handle tool calls, then follow the tutorial that grows one call into a full loop
- GuideFreeOpenAI Function Calling Guide
OpenAI · Intermediate
The same loop on the Responses API: read it after your first provider to see what is universal