https://gcc.gnu.org/bugzilla/show_bug.cgi?id=122992
--- Comment #35 from Frank Ch. Eigler <fche at redhat dot com> ---
> None of the ones I've used so far (ChatGPT, Gemini, Claude) have shown that
> they're competent enough for me to confidently rely on them for technical
> GCC work, or any non-trivial programming for that matter.
As usual, it depends. I'm suggesting y'all give this configuration a try,
precisely because the bunsen MCP provides unique grounding in fresh real
technical data: source code, test logs and results, build logs, bug reports,
almost the whole gamut of our workflow.
For example, asking the followup question:
Has gcc commit 5958e61d12f21cead59e6cdaa592ac648e6f6add been built/tested
yet?
confirms that Andrew's patch corrects the regression, links to the various
builds since this morning where this is shown, and gives a good speculation
about the cause / mechanism).
