[diegosouzapw]

Summary

• Kiro's translator unconditionally injected the literal "(empty)" placeholder into userInputMessage.content whenever a flushed user turn had no accumulated text — even when the turn was a trailing tool-result-only turn (an agentic loop ending in assistant.tool_calls → tool result, with no follow-up user text). Kiro can read that placeholder as if the user actually typed it, disrupting the tool-use loop.
• Fixed by only falling back to "(empty)" when the turn carries no context at all. When the turn already carries toolResults or images, the content is now left as an empty string instead.

Changes

• open-sse/translator/request/openai-to-kiro.ts: flushPending's user branch now computes hasContext = pendingToolResults.length > 0 || pendingImages.length > 0 and only falls back to "(empty)" when hasContext is false (note: this keeps OmniRoute's existing "(empty)" placeholder for genuinely-bare turns — e.g. the assistant-first synthetic filler / session-id hashing logic elsewhere in the file still relies on that literal string; only this one site's unconditional fallback changes).
• tests/unit/translator-openai-to-kiro.test.ts: new regression case builds a payload ending in {role:"assistant", tool_calls:[...]} → {role:"tool", tool_call_id, content} with no trailing user message, and asserts the resulting currentMessage.userInputMessage content is empty (not "(empty)") while userInputMessageContext.toolResults is still populated.

Attribution

Thanks to @jetmiky for the original implementation.

Testing

• node --import tsx/esm --test tests/unit/translator-openai-to-kiro.test.ts — new case FAILs before the fix ('(empty)' !== ''), PASSes after; full 30-test suite green (zero regression, including the "prepends synthetic user when conversation starts with assistant" case that legitimately depends on the "(empty)" placeholder).
• node --import tsx/esm --test tests/unit/translator-ai-sdk-image-parts.test.ts tests/unit/kiro-continue-filler-5231.test.ts — green (adjacent Kiro translator suites, no regression).
• npm run typecheck:core — clean.
• npx eslint open-sse/translator/request/openai-to-kiro.ts tests/unit/translator-openai-to-kiro.test.ts — 0 errors (pre-existing any warnings only).
• npm run check:docs-sync — PASS.

[chatgpt-codex-connector]

You have reached your Codex usage limits for code reviews. You can see your limits in the Codex usage dashboard.

[gemini-code-assist]

Code Review

This pull request prevents the translator from injecting the literal "(empty)" placeholder on trailing tool-result-only turns when converting OpenAI requests to Kiro. Instead, if tool results or images are present, the user content is allowed to remain empty, preventing Kiro from being confused by an artificial placeholder. A corresponding unit test has been added to tests/unit/translator-openai-to-kiro.test.ts to verify this behavior and prevent regressions. No review comments were provided, so there is no additional feedback.

Important

The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.