Shrutam CCA-F Crash › Domain 2 › d2-l05-tool-output-size English Hinglish →

Tool Output Size — Avoiding Context Blow-up

Domain 2 · 20% ~12 min English narration

Scenario anchor

Aaranya IT BFSI team ne ek Claude-powered audit assistant banaya — tool tha jo core banking system se transaction logs fetch karta tha. Production mein pehle hi week, agent ne ek single tool call mein 18,000 rows return kiye. Context window saturate ho gayi, downstream reasoning completely fail hua, aur client escalation aa gayi. Aap us architect ki seat pe hain — samajhna hoga ki tool output size kyun ek first-class design constraint hai, aur kaise aap apne MCP tools ko context-budget-aware banate hain before deployment.

Key Takeaways

Tool output size is a hard context-window budget constraint — every token returned by a tool is deducted from the reasoning budget before the model processes a single word of response.
Design MCP tools with server-side filtering, field projection, and paginated cursors so the tool itself enforces token-aware output contracts, not the caller or the model.
Establish an explicit output-size SLA per tool (e.g., max 2,000 tokens per call) validated at the schema layer — treat it the same way aap apne API gateway mein rate-limit headers enforce karte hain.
Memory anchor: api-response-pagination-contract — agar aapka REST API kabhi unbounded rows return nahi karta, toh aapka MCP tool bhi nahi karega.

← Previous lesson Same lesson in Hinglish Next lesson → Back to overview