REDDIT · REDDIT// 21d agoTUTORIAL

Tool Calling Leaks Into Chat Output

A LocalLLaMA user asks why a model sometimes writes `<tool_call>` into its normal reply instead of emitting a real executable tool call. The thread points to backend/template issues more than raw model behavior, especially in Qwen-style local stacks.

// ANALYSIS

This looks less like “the model forgot how” and more like a serialization bug: if the tool call isn’t preserved as structured metadata, it gets flattened into chat text on the next pass.

–Community replies point to chat template and parser mismatches, especially when using Qwen-family models with local runtimes
–A correct tool-call flow needs separate assistant text and tool-call objects, plus a matching tool-result turn
–If tool-call/result pairs are not replayed exactly, later generations can degrade into plain prose or raw XML tags
–Fixes usually live in the orchestration layer: template alignment, history replay, and strict separation between narration and action
–Once a bad turn poisons the loop, resetting conversation state often “fixes” it temporarily, which is a strong hint the bug is in state handling

// TAGS

llmagentprompt-engineeringautomationopen-sourcetool-calling

DISCOVERED

21d ago

2026-03-21

PUBLISHED

21d ago

2026-03-21

RELEVANCE

8/ 10

AUTHOR

greendude120