Audit Findings → Action Plan

Before you start, ask me about each [input] below — one question at a time, with ready-made answers to pick from; skip whatever you can already tell.

Turn the open audit findings of [client: if not listed, call roasr_list_clients and let me pick] into a prioritised plan for the week, sized by how much spend each problem touches.

1. **Load the RoASr method.** Call roasr_load_skill('roasr_frame'), roasr_load_skill('audit_scoring') and roasr_load_skill('report_format') and follow them exactly.
2. **Frame the data.** Call roasr_limits, roasr_list_clients and roasr_list_accounts: plan, niche, margin, KPI, account currencies, sync status and dataSince. Size impact over [spend window: last 30 days (default) | last 14 days | last 60 days | custom]. If a window starts before dataSince, say so and narrow it. If currency.mixed is true, never sum money across accounts; rank within each account. An account with sync failed or never gets no conclusions.
3. **Collect.** roasr_list_findings for this client. Keep [scope: all open findings (default) | high severity only | high and medium]. Ignore resolved and dismissed ones; a finding keeps its identity between audits, so mention how long it has stayed open if the data shows it.
4. **Group.** Sort findings by severity, then by the category of the audit_scoring checklist they belong to. Merge duplicates that point to the same root problem and say so.
5. **Size.** roasr_metrics_campaigns at campaign level (adset level when a finding is about ad sets), sorted by spend. Tie each finding to the campaigns it touches and show their share of the account's spend in the window. Findings about the whole account take the whole account's spend. Show "spend at stake", not "savings": do not promise a number the data cannot support. Classify campaigns by their settings and data, not by their names.
6. **Plan the week.** 5–8 steps ordered by severity × share of spend ÷ effort. For each: what to do, in which account and campaign, how to verify it worked in the next data pull, and who decides. Put quick, safe fixes first; leave anything that restarts learning for a separate, labelled step.

**Output:** open with one sentence: how many findings are open, how much spend they touch and the one thing to do first. Then a table: severity · category · finding · account/campaigns · spend at stake · effort. Then the week plan as 5–8 steps grouped High / Medium / Nice-to-have, each with the expected effect.

**Output rules:**
- Recommendations only — never apply changes; RoASr tools are read-only.
- Do not add up spend of overlapping findings: one campaign counts once in the total.
- With no margin set for the client, talk ROAS and CPA, not profit.
- If there are no open findings, say so and suggest a full audit instead of making work up.
- Answer in the language of this prompt.