Best possible solution:

Land the local helper substitution after contributor-supplied real-behavior proof is added, ideally with a focused formatError JSON-fallback regression; no core or plugin API change is needed.

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

Yes at source level. Current main uses raw UTF-16 code-unit slicing in safeStringify, so a JSON-stringified fallback error whose high surrogate lands at the 800-code-unit boundary can return a dangling surrogate before the ellipsis; I did not run a live WhatsApp session.

Is this the best way to solve the issue?

Yes. Reusing the existing truncateUtf16Safe SDK helper at the local truncation point is the narrow maintainable fix, with the remaining gap being proof and preferably a focused call-site regression.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against 34d257713ea3.

Label changes:

• add P3: This is a very small WhatsApp diagnostic/error-string correctness fix with limited blast radius and no evidence of a broken core workflow.
• add rating: 🧂 unranked krab: Overall readiness is 🧂 unranked krab; proof is 🧂 unranked krab and patch quality is 🐚 platinum hermit.
• add status: 📣 needs proof: The PR needs real behavior proof before ClawSweeper can clear the contributor ask. Needs real behavior proof before merge: Missing: the PR body gives only mechanical evidence; the contributor should add redacted terminal output, copied live output, runtime logs, or a validation transcript showing after-fix WhatsApp/session error formatting, then update the PR body to trigger re-review.

Label justifications:

• P3: This is a very small WhatsApp diagnostic/error-string correctness fix with limited blast radius and no evidence of a broken core workflow.
• rating: 🧂 unranked krab: Overall readiness is 🧂 unranked krab; proof is 🧂 unranked krab 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: Missing: the PR body gives only mechanical evidence; the contributor should add redacted terminal output, copied live output, runtime logs, or a validation transcript showing after-fix WhatsApp/session error formatting, then update the PR body to trigger re-review.

PR surface:

Source +2. Total +2 across 1 file.

What I checked:

• Repository policy read and applied: The full root policy and scoped extensions policy were read; the review applied the plugin-boundary rule that extension production code should use openclaw/plugin-sdk/* surfaces and the deep-review requirement for PR verdicts. (AGENTS.md:1, 34d257713ea3)
• PR diff is focused on the WhatsApp formatter: The proposed head adds the SDK helper import and replaces raw.slice(0, limit) with truncateUtf16Safe(raw, limit) in the safeStringify truncation branch. (extensions/whatsapp/src/session-errors.ts:29, 240a8b74dba7)
• Current main still has the unsafe truncation site: On current main, safeStringify still truncates the JSON string with raw slice(0, limit) before appending the ellipsis. (extensions/whatsapp/src/session-errors.ts:30, 34d257713ea3)
• Helper contract matches the intended fix: truncateUtf16Safe clamps/floors the limit and delegates to sliceUtf16Safe, which avoids returning dangling surrogate halves at slice edges. (packages/normalization-core/src/utf16-slice.ts:44, 34d257713ea3)
• SDK subpath is an existing extension-safe import: text-utility-runtime already exports sliceUtf16Safe and truncateUtf16Safe, and the root package exports the openclaw/plugin-sdk/text-utility-runtime subpath. (src/plugin-sdk/text-utility-runtime.ts:21, 34d257713ea3)
• Sibling WhatsApp code already uses the helper: WhatsApp auto-reply utilities and echo logging already import and use truncateUtf16Safe, so the PR follows an established local pattern rather than adding a new abstraction. (extensions/whatsapp/src/auto-reply/util.ts:3, 34d257713ea3)

Likely related people:

• steipete: GitHub path history shows steipete on the WhatsApp session-errors history and normalization-core extraction work, and PR discussion for related UTF-16 truncation work shows steipete handled land-ready validation and merge evidence. (role: recent area contributor and reviewer; confidence: high; commits: 30a94dfd3b7c, 58912f8fd842, 062f88e3e3af; files: extensions/whatsapp/src/session-errors.ts, extensions/whatsapp/src/session.test.ts, packages/normalization-core/src/utf16-slice.ts)
• obviyus: Recent GitHub path history for src/plugin-sdk/text-utility-runtime.ts shows obviyus working on the plugin SDK utility export surface that this PR imports from. (role: recent plugin SDK surface contributor; confidence: medium; commits: d498b1cce49c, 210877a73e25; files: src/plugin-sdk/text-utility-runtime.ts)
• RomneyDa: Recent WhatsApp session test history includes RomneyDa's WebSocket URL and session-surface work, making them a plausible adjacent routing candidate for WhatsApp runtime validation. (role: recent adjacent WhatsApp contributor; confidence: medium; commits: 2ec670898018, 0deddfa20428; files: extensions/whatsapp/src/session.test.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.