Continue + Ollama (qwen2.5-coder) in agent mode: replies come back as raw JSON tool calls instead of running tools

Context

Source(s): https://github.com/continuedev/continue/issues/7085 Status: source reports fixed; not reproduced by us (issue closed as stale; fix confirmed by several commenters).

Problem

In Continue's agent mode with a local Ollama model, the chat shows JSON blobs like {"name": ..., "arguments": ...} and no tool ever runs.

Error (redacted / normalized)

Chat prints {"name": ..., "arguments": ...} and no tool runs

Applies when

Continue VS Code extension, agent mode, local Ollama qwen2.5-coder 7b/14b/32b.

Confirm it is this problem

Agent-mode replies contain the tool-call JSON as plain text and nothing executes.

Not this when / boundary

One commenter said that with this setting on, tool calls work in Continue, but other agent CLIs (qwen-code, continue cli, opencode) with the same model still printed raw JSON - the setting only changes Continue. (our note: models with solid native tool calling don't need this.)

Root cause

A Continue contributor said the model doesn't do native tool calling reliably. The "system message tools" setting describes the tools in the system prompt and parses the calls out of the text instead.

Fix

1. In Continue's settings, turn on "Only use system message tools".
2. Restart VS Code (one confirmer restarted "all the things"). → expect: the same prompt runs the tool.

How to verify

The same prompt now runs the tool instead of printing JSON.

What did not work

Switching quantization (q8_0). One user on Linux Mint got an empty Python file after turning the setting on.

Risks

The continuedev/continue repository README says it is no longer actively maintained and is read-only.

Verified scope

Continue + Ollama qwen2.5-coder, checked Oct 2026