Best possible solution:

Land the latest head after maintainer review and green checks if the opt-in draft --reply-all contract is approved; otherwise keep the linked issue as the behavior-policy discussion.

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

Yes. Source inspection shows current main prepares draft reply metadata but builds draft recipients only from explicit To/Cc/Bcc while allowing missing To, matching the linked report.

Is this the best way to solve the issue?

Yes for the latest PR shape, subject to maintainer approval of the new CLI flag. The explicit opt-in mode preserves no-recipient drafts by default while matching the existing send/reply-all recipient derivation model.

AGENTS.md: found and applied where relevant.

Codex review notes: model internal, reasoning high; reviewed against 23b847ca53cf.

Label justifications:

• P2: This is a normal-priority Gmail draft reply workflow fix with clear scripted-use impact and limited blast radius.
• merge-risk: 🚨 message-delivery: The PR changes who appears in To/Cc/Bcc for reply-shaped drafts when --reply-all is used, which is message-delivery-sensitive.
• rating: 🐚 platinum hermit: Overall readiness is 🐚 platinum hermit; proof is 🌊 off-meta tidepool and patch quality is 🐚 platinum hermit.
• status: 👀 ready for maintainer look: ClawSweeper has no concrete contributor-facing blocker left for this PR. Not applicable: This is an autogenerated ClawSweeper bot PR, so the external-contributor real behavior proof gate does not apply; the PR body reports targeted tests and make ci.

What I checked:

• Repository policy read: AGENTS.md was read fully; its PR review-mode guidance was applied by inspecting the PR with gh without changing branches or files. (AGENTS.md:31, 23b847ca53cf)
• Current main lacks draft reply-all: Current main exposes reply target and quote fields for drafts, but no ReplyAll field, and buildDraftMessage passes only explicit To/Cc/Bcc values to the MIME builder. (internal/cmd/gmail_drafts.go:254, 23b847ca53cf)
• Current main allows no-recipient drafts: The MIME builder only rejects missing To when AllowMissingTo is false, matching the existing draft behavior that this PR preserves by making recipient inference opt-in. (internal/mailmime/mime.go:79, 23b847ca53cf)
• Existing send path has reply-all parity: gog gmail send already validates --reply-all, derives recipients from the reply target, and lets explicit recipient flags override inferred fields. (internal/cmd/gmail_send.go:82, 23b847ca53cf)
• PR implementation is bounded: The PR diff adds draft ReplyAll, validates that it has a reply target, calls shared reply-recipient logic, and preserves explicit recipient replacement semantics. (internal/cmd/gmail_drafts.go:262, c89a054abf91)
• Focused regression coverage added: The PR adds tests for draft create inference, missing-target validation, explicit recipient replacement, and draft update inference. (internal/cmd/gmail_drafts_cmd_test.go:1426, c89a054abf91)

Likely related people:

• steipete: Current blame for the draft compose path points to the v0.27.0 release authored by Peter Steinberger, and merged reply-command work added the shared reply-recipient surface this PR reuses. (role: recent Gmail write-path contributor; confidence: high; commits: 274a2a61da98, ba350d6824ce; files: internal/cmd/gmail_drafts.go, internal/cmd/gmail_recipients.go, internal/cmd/gmail_reply_commands.go)
• antons: Merged PR history for allowing drafts without recipients touched the same draft and MIME validation behavior that constrains this fix. (role: draft no-recipient behavior introducer; confidence: medium; commits: 471324b4511a; files: internal/cmd/gmail_drafts.go, internal/cmd/gmail_drafts_cmd_test.go, internal/cmd/gmail_mime.go)
• chrischall: Related review context traces recent draft thread-ID and attachment-preservation work to this author on the same draft create/update surface. (role: recent draft reply parity contributor; confidence: medium; commits: 22fbc35c7c18, 3d5c9cef65fa; files: internal/cmd/gmail_drafts.go, internal/cmd/gmail_drafts_cmd_test.go)
• salmonumbrella: Related review context traces the original lower-level gmail send --reply-all recipient-routing behavior to this contributor, which is the parity model for the draft change. (role: adjacent recipient-routing contributor; confidence: medium; commits: bf7d500849a0; files: internal/cmd/gmail_send.go)

• 🦀 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.