Prepare a bounded Apple Ads search experiment with an explicit spend gate

[lightcloud00]

Outcome

Prepare—but do not activate—a small search-results acquisition test whose economics and stop rules are known in advance.

Dependencies

• Public App Store version in Create signed archive and run TestFlight beta go/no-go #9
• Organic acquisition/revenue baseline in Establish the privacy-preserving launch revenue scorecard #14
• Intent-specific custom product pages in Create intent-specific App Store custom product pages #17
• Current pricing decision in Review the $9.99 Lifetime Pro price from real proceeds and conversion data #20

Preparation

• Build Brand, Find My Car, Meter Reminder, and Garage intent groups with exact/broad negatives documented.
• Match each non-brand group to the corresponding custom product page.
• Calculate maximum sustainable paid-install cost from actual proceeds per download; use a more conservative cap for the pilot.
• Specify countries, daily budget, total time-box, end date, CPA cap, and hard loss limit.
• Define stop rules for irrelevant queries, poor tap-through, weak install conversion, or acquisition cost above the cap.
• Prepare a one-page approval packet with worst-case spend.

Protected action

Creating or activating a campaign spends money and is not authorized by this issue. Obtain explicit owner approval for the exact account, countries, dates, daily budget, worst-case total, and CPA cap.

Acceptance

• Dry-run campaign structure and approval packet are committed.
• Owner approval is linked before activation.
• Post-test report records spend, taps, installs, purchases/proceeds where attributable, and a continue/stop decision.

Apple Ads guidance: https://ads.apple.com/app-store/help/campaigns/0005-create-campaigns

---

Scope / non-scope

Only the work and evidence already described in this issue are in scope. Merge, deployment, provider or store mutation, spend, billing, customer contact, device acceptance, indexing, and release remain separate gates unless this issue explicitly includes them and fresh proof is recorded.

Plan

1. Revalidate the current evidence above against exact current source or live state.
2. Complete the scoped work and unchecked items above in dependency order, preserving every non-goal and stop gate.
3. Run the focused tests and proof already named; add regression coverage for any corrected defect.
4. Record the exact revision, receipt, and applicable live readback before closure.

Evidence and closure

Close only from exact current-source evidence plus the applicable independent readback. Keep local, GitHub/CI, provider, deployment, device, store, commerce, indexing, traffic, and revenue evidence distinct; one does not imply another.

Owner / lane

The repository delivery lane owns ordinary implementation and verification. The operator owns only actions explicitly marked as operator-only or protected.

Resolution plan

• Verified: 2026-08-25T05:34:24Z against main@04a6cd5e808720983471ee9595fd40bc94e1f0fe (commit time 2026-08-25T00:53:29Z).
• Source basis: Original issue body, all 0 comment(s), current default-branch tree, disposition-appropriate source-path checks, and referenced pull-request states.
• Disposition: Operator or external gate - execute only the bounded proof steps and stop at protected actions.
• Outcome: Prepare—but do not activate—a small search-results acquisition test whose economics and stop rules are known in advance.

Current evidence and corrected premise

The issue remains open and requires the resolution steps below.
All 0 issue comment(s) were read. There are no comments.

Resolution steps

1. Refresh the authoritative external state and bind it to the exact account, app, build, deployment, or provider identity.
2. Reconcile that readback with the issue-authored work below and identify the smallest remaining action.
3. Prepare the protected action, rollback, and failure receipt without exposing credentials or inferring authority.
4. Stop for the required operator decision or action, then execute it once through the owning workflow.
5. Validate the official readback and map it to every acceptance item before closure.

Grounded issue-authored detail

Scope / non-scope

Only the work and evidence already described in this issue are in scope. Merge, deployment, provider or store mutation, spend, billing, customer contact, device acceptance, indexing, and release remain separate gates unless this issue explicitly includes them and fresh proof is recorded.

Tests and acceptance

Acceptance

• Dry-run campaign structure and approval packet are committed.
• Owner approval is linked before activation.
• Post-test report records spend, taps, installs, purchases/proceeds where attributable, and a continue/stop decision.

Apple Ads guidance: https://ads.apple.com/app-store/help/campaigns/0005-create-campaigns

---

Dependencies and stop gates

Dependencies

• Public App Store version in Create signed archive and run TestFlight beta go/no-go #9
• Organic acquisition/revenue baseline in Establish the privacy-preserving launch revenue scorecard #14
• Intent-specific custom product pages in Create intent-specific App Store custom product pages #17
• Current pricing decision in Review the $9.99 Lifetime Pro price from real proceeds and conversion data #20
• Do not close this issue until every acceptance item is mapped to exact merged-head evidence.

Next AI first action

Only the work and evidence already described in this issue are in scope. Merge, deployment, provider or store mutation, spend, billing, customer contact, device acceptance, indexing, and release remain separate gates unless this issue explicitly includes them and fresh proof is recorded.

[lightcloud00]

Verified issue disposition — blocked_external

This is the current managed disposition for lightcloud00/parknudge#22 at default-branch 5c3d185143a2.

Current evidence

• Current stable snapshot for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 at 5c3d185143a2 keeps the issue open with body hash cf6b67adf55a; the validated grounded planner classified it as blocked_external.

Files or owner actions

• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22, bind the issue title 'Prepare a bounded Apple Ads search experiment with an explicit spend gate' to one existing source path or named owner action before implementation.

Bounded plan

• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: Resolve the recorded targeted-source diagnostic or bind a canonical checkout before any implementation attempt.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: Complete every read-only and local source check that does not require the external owner or protected system.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: Bind the remaining decision to the exact issue surfaces: Prepare, bounded, Apple, search, experiment, explicit, spend, gate.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: Record the required owner input as one explicit question with a recommended default and impact of deferral.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: Continue the portfolio queue; resume this issue only when the named external proof or decision arrives.

Commands

• Run git grep -n -F Prepare main for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 and retain sanitized output.
• Run git grep -n -F bounded main for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 and retain sanitized output.
• Run git grep -n -F Apple main for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 and retain sanitized output.
• Run git grep -n -F search main for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 and retain sanitized output.
• Run git grep -n -F experiment main for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 and retain sanitized output.
• Run python3 -m pytest -q for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 and retain sanitized output.

Acceptance tests

• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: [ ] Dry-run campaign structure and approval packet are committed.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: [ ] Owner approval is linked before activation.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: [ ] Post-test report records spend, taps, installs, purchases/proceeds where attributable, and a continue/stop decision.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: [ ] Dry-run campaign structure and approval packet are committed.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: [ ] Owner approval is linked before activation.
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: [ ] Post-test report records spend, taps, installs, purchases/proceeds where attributable, and a continue/stop decision.

Risks

• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22, a false source binding could widen the change or invalidate this plan; Revert only the exact task-branch change after capturing source and runtime identity; never reset or restore an owner worktree.

Stop gates

• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: no issue close, reopen, body edit, pull-request, branch, or merge mutation while implementing this plan
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: the managed plan comment and disposition-specific readiness labels are the only authorized issue mutations in the planning lane
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: no push, protected-branch update, force-push, or merge
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: no deploy, upload, App Store Connect, TestFlight, or public release
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: no credential value disclosure, rotation, billing, or provider change
• For Prepare a bounded Apple Ads search experiment with an explicit spend gate #22: no owner-worktree restore, reset, checkout, process signal, or queue reorder

Named owner action

• Gus must complete or coordinate the protected action 'Prepare a bounded Apple Ads search experiment with an explicit spend gate' for Prepare a bounded Apple Ads search experiment with an explicit spend gate #22, then attach sanitized evidence at proof/parknudge-22/acceptance.json.

Proof target

• proof/parknudge-22/acceptance.json

Retry trigger

• Retry Prepare a bounded Apple Ads search experiment with an explicit spend gate #22 only after proof/parknudge-22/acceptance.json records the title-specific external action as complete.

---

Ledger schema: aos.github-issue-plan-coverage.v1 · issue body: cf6b67adf55a