feat: surface model content-filter refusals as a blocked chat error - #27118

When a provider's real-time safety classifiers block a response
(e.g. Anthropic stop_reason "refusal"), the turn previously ended
silently: the fantasy SDK mapped the stop reason to unknown and the
empty-content finish looked like a normal stop, leaving the UI stuck
on "Thinking" with no output.

Bump the fantasy fork to pick up refusal -> FinishReasonContentFilter
mapping with stop_details carried as RefusalMetadata (fork change 15,
coder/fantasy#41). chatloop now turns an empty content-filter finish
into a terminal, non-retryable content_filter classified error, using
the provider's explanation as the message and the refusal category as
the detail, with a generic per-provider fallback message. The UI
renders it as "Response blocked".

Partial-content refusals (classifier blocks after some output) still
persist the partial text and end the turn normally; surfacing those is
a separate product decision.

Coder Agents generated.

…entFilterError

Merge TestContentFilterMessage into TestContentFilterError as a
table-driven test. contentFilterError routes through ContentFilterMessage
for both the metadata and default-fallback paths, so asserting the
classified message covers the verbatim category, default, whitespace,
and unknown-provider cases without a separate test.