Best possible solution:

Keep the focused shared-normalizer fix and merge it after redacted real behavior proof plus refreshed exact-head checks confirm the Parallel provider path works.

Do we have a high-confidence way to reproduce the issue?

Yes. A current-main Node check reproduces the raw-slice surrogate split, and current source uses that same raw slice in the Parallel normalizer.

Is this the best way to solve the issue?

Yes. Using the existing shared UTF-16-safe helper at the shared Parallel normalizer is the narrowest maintainable fix; schema-only or transport-specific changes would leave runtime duplication or one provider path unfixed.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against e591dcfa8213.

Label justifications:

• P2: The PR fixes a bounded UTF-16 normalization defect in the Parallel web-search provider with limited blast radius to non-BMP text near field limits.
• rating: 🦪 silver shellfish: Overall readiness is 🦪 silver shellfish; proof is 🦪 silver shellfish and patch quality is 🐚 platinum hermit.
• status: 📣 needs proof: The PR needs real behavior proof before ClawSweeper can clear the contributor ask. Needs real behavior proof before merge: Only unit/static validation is present; the contributor should add redacted real Parallel provider proof such as terminal output, logs, or a request transcript after the fix. After adding proof, update the PR body; ClawSweeper should re-review automatically. If it does not, the PR author or someone with repository write access can comment @clawsweeper re-review.

PR surface:

Source +1, Tests +16. Total +17 across 2 files.

What I checked:

• Repository policy read: Root and scoped extension AGENTS.md were read fully; the extension policy allows production imports from openclaw/plugin-sdk/*, which matches this PR's new helper import. (extensions/AGENTS.md:1, e591dcfa8213)
• Current main failure path: Current main still truncates Parallel objective, client_model, and search_queries with raw String.slice calls, so a surrogate pair can be cut at the configured field boundary before this PR. (extensions/parallel/src/parallel-search-normalize.ts:59, e591dcfa8213)
• Focused reproduction: A current-main Node check using the same 200-code-unit raw slice returned a string with a lone high surrogate for an emoji crossing the cap. (e591dcfa8213)
• Shared caller coverage: Both paid Parallel REST and free Parallel MCP runtimes call the same normalization functions, so changing the shared normalizer covers both entry points without transport-specific duplication. (extensions/parallel/src/parallel-web-search-provider.runtime.ts:197, e591dcfa8213)
• Helper contract: truncateUtf16Safe floors the limit and delegates to sliceUtf16Safe, whose implementation backs off when a cut would leave a dangling surrogate half. (packages/normalization-core/src/utf16-slice.ts:43, e591dcfa8213)
• PR head implementation: At the reviewed PR head, the three raw truncations are replaced with truncateUtf16Safe while preserving the existing limits and request shape. (extensions/parallel/src/parallel-search-normalize.ts:60, a6fe951e229f)

Likely related people:

• steipete: Git blame points the current Parallel normalizer and shared UTF-16 helper export to commit 6e792f0, whose live merged PR metadata identifies steipete as author and merger; the path history is otherwise thin in this checkout. (role: introduced current implementation / recent area contributor; confidence: medium; commits: 6e792f08f88f; files: extensions/parallel/src/parallel-search-normalize.ts, packages/normalization-core/src/utf16-slice.ts, src/plugin-sdk/text-utility-runtime.ts)

• 🦀 challenger crab: rare, exceptional readiness with strong proof, clean implementation, and convincing validation.
• 🦞 diamond lobster: very strong readiness with only minor maintainer review expected.
• 🐚 platinum hermit: good normal PR, likely mergeable with ordinary maintainer review.
• 🦐 gold shrimp: useful signal, but proof or patch confidence is still limited.
• 🦪 silver shellfish: thin signal; proof, validation, or implementation needs work.
• 🧂 unranked krab: not merge-ready because proof is missing/unusable or there are serious correctness or safety concerns.
• 🌊 off-meta tidepool: rating does not apply to this item.

Shiny media proof means a screenshot, video, or linked artifact directly shows the changed behavior. Runtime, network, CSP, and security claims still need visible diagnostics.

• ClawSweeper keeps one durable marker-backed review comment per issue or PR.
• Re-runs edit this comment so the latest verdict, findings, and automation markers stay together instead of adding duplicate bot comments.
• A fresh review can be triggered by eligible @clawsweeper re-review comments, exact-item GitHub events, scheduled/background review runs, or manual workflow dispatch.
• PR/issue authors and users with repository write access can comment @clawsweeper re-review or @clawsweeper re-run on an open PR or issue to request a fresh review only.
• Maintainers can also comment @clawsweeper review to request a fresh review only.
• Fresh-review commands do not start repair, autofix, rebase, CI repair, or automerge.
• Maintainer-only repair and merge flows require explicit commands such as @clawsweeper autofix, @clawsweeper automerge, @clawsweeper fix ci, or @clawsweeper address review.
• Maintainers can comment @clawsweeper explain to ask for more context, or @clawsweeper stop to stop active automation.

• reviewed 2026-07-07T13:27:02.566Z sha a6fe951 :: needs real behavior proof before merge. :: none