[pacocartones]

Summary

• The Claude to OpenAI stream translator and the Bedrock executor turned every stop reason they did not know into stop. A Claude refusal, a Bedrock content_filtered or guardrail_intervened stop and a model_context_window_exceeded stop all reached OpenAI-format clients as a clean completion.
• They now map to content_filter and length, the same values utils/finishReason.ts already uses for these reasons.

Related Issues

• Related to fix(sse): map Responses response.incomplete to length/content_filter finish reasons (#15489) #15867 (same mapping for Responses incomplete)

Target Branch

• Base is the active release/v3.8.x branch (the highest open release/v*), not main
• New feature? From 3.8.55 on it is held with the v4-feature label for the v4 channel

Validation

• Change type: other (translator / executor)
• Focused tests and category gates from the golden path
• npm run lint (eslint on the changed files: clean)
• Reconciled with the current active release base; focused checks rerun afterward
• Production-code changes include a new or updated automated test in this PR

Tests Added Or Updated

• tests/unit/finish-reason-refusal-mapping.test.ts (new): 9 tests. 6 fail on the base and all 9 pass with the fix.

Coverage Notes

• open-sse/translator/response/claude-to-openai.ts and open-sse/executors/bedrock.ts: the new test drives both the streaming and the non-streaming Bedrock paths, and checks that the existing mappings stay as they were.

Reviewer Notes

• 55 test files that import these modules pass, except one inherited failure that also fails on the base: provider-validation-specialty.test.ts › gemini-web validator.

⚠️ base-red inherited: #15306