ad-complianceReactive Meta policy engine — diagnoses a rejection, proposes a fix, and (only when explicitly requested) pre-screens ad copy in Audit mode.

=== AGENT RUN OPEN ===
agent_name:       ad-compliance
triggered_by:     operator
expert_namespace: ida
ad_id:            120215493827640000
rejection_reason: "This ad can't run because it implies personal attributes about people who see it"
started_at:       2026-06-11T09:14:22Z
status:           running
run_id:           acr-20260611-0914

=== STEP 8A: REJECTION REASON DIAGNOSIS ===

Rejection reason: "This ad can't run because it implies personal attributes..."
Decision tree mapping: PERSONAL ATTRIBUTES (Step 2 — #1 rejection reason)
Confidence: HIGH

The rejection reason maps cleanly to Meta's personal-attributes category. This is the most
common rejection in the IDA niche. Running full scan now to locate every violation.

=== STEPS 1–6: FULL RE-SCAN ===

[Plain code scan output — bounded shortlist for model reasoning]

Step 1 — Prohibited Content:        CLEAR (all 13 categories)
Step 3 — Restricted Content:        COMPLIANT — income/business niche, no Special Ad Category required
Step 5 — Landing Page (LP):         CLEAR (HTTPS confirmed, earnings disclaimer present at footer,
                                    "results not typical" near testimonials, no FB logo found)
Step 6 — Before/After:              N/A (no transformation images in this creative)

Step 2 — Personal Attributes:       2 VIOLATIONS FOUND
  [V1] Primary text, line 1:
       "Are you still trading your time for $3K design projects when your customers
        have $30K budgets?"
       Rule triggered: "you" + negative financial state ("still trading time for $3K")
       — asserts the viewer is in a financially inferior position. Tier: instant rejection.

  [V2] Primary text, line 4:
       "If you're struggling to raise your fees, here's why..."
       Rule triggered: "If you're struggling with [X]" — banned pattern regardless of niche.
       Tier: instant rejection.

Step 4 — Banned Words:              TIER 2 — 1 item
  [B1] Primary text, line 3:
       "guaranteed income growth" — Tier 2 banned phrase (income claims).
       Instant rejection.

=== ROOT CAUSE SUMMARY ===

Primary root cause:  Personal Attributes — V1 and V2 above (HIGH confidence)
Secondary violation: Tier 2 income claim — B1 above
Total blocking violations: 3
Verdict: FIX BEFORE SUBMITTING (do not appeal without fixing first)

=== FIX PROTOCOL (STEP 8B) ===

Fix F1 — resolves V1
  Original:  "Are you still trading your time for $3K design projects when your customers
              have $30K budgets?"
  Technique: Aspiration Framing (forward-looking, removes viewer-negative assertion)
  Revised:   "What would change if your next interior design project paid $30K instead
              of $3K?"
  Why:       Removes "you" + negative state. The revised version invites curiosity
             (aspiration) without asserting the viewer is currently underpaid.

Fix F2 — resolves V2
  Original:  "If you're struggling to raise your fees, here's why..."
  Technique: Third-Person Story (shifts struggle to a named character, not the viewer)
  Revised:   "Amelia charged $3K for years before she understood what was holding
              her pricing back."
  Why:       Completely removes second-person address. Amelia's story creates the
             same emotional bridge without asserting anything about the viewer.

Fix F3 — resolves B1
  Original:  "guaranteed income growth"
  Revised:   "a path to consistent income growth"
  Why:       "guaranteed" is a Tier 2 banned word. "a path to" removes the guarantee
             while preserving the promise of direction and momentum.

=== APPEAL DECISION ===
Appeal: NO
Reason: All three violations are genuine policy hits — not false positives. Fix first.
Create a NEW ad using the revised copy above. Do not edit the rejected ad in place
(in-place editing triggers a re-review flag on the account).

=== WAIT RECOMMENDATION ===
Wait period: 24 hours (standard recommendation; escalate to 48h if account
has prior violation history this month).

=== FINAL CONFIRMATION ===
"Diagnostic complete. Root cause: Personal Attributes (HIGH confidence) +
Tier 2 income claim. 3 violations fixed. Create a new ad with the revised copy
above. Wait 24h before submitting. No appeal recommended."

=== AGENT RUN CLOSE ===
status:        success
status_detail: 3 violations found (V1 personal-attribute/aspiration, V2 personal-attribute/
               third-person, B1 Tier2-income). All fixed. No appeal. New ad recommended.
               Models used: Haiku 4.5 (Meta MCP ingest), Opus 4.8 (diagnosis + rewrite).
               OpenRouter cost: logged to run_id acr-20260611-0914.