#105)
ResponseTooLargeError's behavior was spelled out near-verbatim in
docs/errors.md, architecture/errors.md, and architecture/client.md;
architecture/errors.md now cross-references the other two instead of
restating them.
The "why not respx" paragraph was duplicated between docs/testing.md
and architecture/testing.md, and the prior audit-fix pass (#103) had
mechanically changed a "breaks across httpx major versions" claim to
say httpx2 instead. Verified against respx's own README (requires
httpx 0.25+, no stated httpx2 support) and GitHub history: the
breakage claim is real but about httpx, not httpx2. Corrected the
claim and deduped, keeping the full rationale in docs/testing.md.
Summary
Test plan