Goose 1.6.0: "Error: Unsupported operation: streaming not implemented" with Ollama
In Goose 1.6.0, a prompt failed with
Error: Unsupported operation: streaming not implementedand the session was interrupted before the model replied. The report appeared in a feature request asking for streaming support in the Goose Ollama integration (ollama.rs). Is this a known limitation, and is there a fix or a version that includes streaming?Originally asked on GitHub: https://github.com/aaif-goose/goose/issues/3527
3 answers
Status: closed as COMPLETED \u2014 the streaming feature request was accepted as done, but the thread does not name the release that contains it. Goose 1.6.0 clearly does not include the change.
This thread is a feature request rather than a bug investigation. A user on Goose 1.6.0 reproduced the failure: after entering
hello?, Goose printedHashing hope hypotheses...and then failed withError: Unsupported operation: streaming not implemented, interrupted the session, and removed the last message. No root cause beyond the error text was identified, and no configuration or workaround was discussed.The requested implementation, described in the issue body, was to send
stream: truein the Ollama API request and modify thecompletefunction inollama.rsto consume the server-sent events (SSE) stream, rendering chunks as they arrive. The issue being closed as COMPLETED indicates the maintainers accepted that work, but nobody in the thread confirms streaming behavior on a newer build.Practical steps:
- Upgrade Goose to a release newer than 1.6.0, the version where the error was reproduced. As general advice, check the release notes for a mention of Ollama streaming to confirm your build contains the change.
- No flag, config value, or workaround that enables streaming on older versions was given in the thread.
Did this work for you?What a GitHub user tried on GitHub (Outcome unclear)
The issue author proposed implementing streaming by adding
stream: trueto the Ollama API request and rewriting thecompletefunction inollama.rsto process SSE chunks incrementally. The thread contains no report that this approach was tested; the issue was later closed as COMPLETED.Source: https://github.com/aaif-goose/goose/issues/3527 (opens external site)
Did this work for you?What @anthonyrawlins tried on GitHub (Didn't work)
A user on Goose 1.6.0 entered a prompt and got
Error: Unsupported operation: streaming not implemented; Goose interrupted the session and removed the last message.Source: https://github.com/aaif-goose/goose/issues/3527#issuecomment-3216076638 (opens external site)
Did this work for you?
Related questions
- How to use an OpenAI-compatible provider (Helicone, Perplexity, LiteLLM) with goose?
- Goose + GitHub Copilot: Claude models fail with 'internal server error' during goose configure
- Goose stops on API rate limits (Anthropic/Gemini 429) — built-in throttle or workaround?
- Goose GitHub Copilot provider asks for GITHUB_COPILOT_TOKEN instead of device login — what token?
- Goose + Gemini 400 Bad Request with MCP extensions: 'only enum and date-time are supported for STRING type'