A public-health portal has a vague analyze_content tool that overlaps lab extraction. What redesign best reduces selection errors?
Select an answer to reveal the explanation.
Short Explanation
Vague analyze_content is a trap next to lab work. Rename it extract_lab_results and write a lab-specific description so the health portal stops mis-firing tools.
Full Explanation
A public-health portal with a vague analyze_content tool that overlaps lab extraction invites selection errors. The best redesign is to rename it to extract_lab_results and rewrite the description for lab-specific use—inputs (specimen or report IDs), outputs (structured lab fields), and boundaries (not general document summary).
Purpose-specific naming and description work because they collapse functional overlap. When the contract clearly means lab extraction, the health agent stops treating a generic analyzer as interchangeable with true lab tools, improving accuracy for outbreak reporting and clinician-facing portals.
Keeping the vague name and adding three more equally vague analysis tools multiplies collision and makes selection worse. Deleting all lab tools and asking Claude to guess results from chat alone removes authoritative MCP evidence and risks fabricated lab values. Hiding the tool from the schema while still expecting the agent to call it breaks discoverability; models cannot reliably invoke tools that are not exposed in the tool list.
Exam caveat: rename and rewrite together—cosmetic rename without a lab-specific description often leaves residual overlap. Operational check: after the rename/rewrite, run portal prompts that need lab extraction versus general content analysis and confirm only extract_lab_results is selected for lab intents.