Ads Compliance Review

Review ads and campaign setup for policy compliance on Meta (Facebook/Instagram), Google Ads, TikTok Ads, YouTube Ads, Amazon Ads, and DSP/programmatic.

Security Scan

The skill's instructions, scope, and requirements are consistent with an ads compliance review tool: it is instruction-only, requests no credentials or installs, and stays within its stated purpose.

Purpose & Capability

Name and description match the SKILL.md: policy screening, violation detection, and compliant rewrites across the listed ad platforms. Nothing in the files asks for unrelated capabilities (no cloud credentials, no system access).

Instruction Scope

Runtime instructions focus on normal compliance review tasks (normalize inputs, detect policy issues, produce findings and handoff payloads). The SKILL.md does not instruct the agent to read local files, query external endpoints, access environment variables, or exfiltrate data beyond what the user supplies.

Install Mechanism

No install spec and no code files — the skill is instruction-only, so nothing will be written to disk or downloaded during install.

Credentials

No environment variables, credentials, or config paths are requested. The absence of required secrets is proportionate to the described review functionality.

Persistence & Privilege

Skill is not always-enabled and does not request persistent/system-wide privileges. It does not modify other skills or system configuration.

Assessment

This skill appears internally consistent and safe as an instruction-only compliance reviewer. Before using it, avoid pasting real account credentials or tokens into prompts — provide only the campaign data needed (ad text, landing page URLs, targeting criteria, and performance metrics). If you intend to have the agent perform account-level actions (publish, pause campaigns, or connect ad accounts), require an explicit, separate integration that uses platform APIs and vetted credentials; verify any automated changes manually before applying them to live campaigns.

Purpose

Detect compliance risks and provide concrete compliant rewrites before publishing.

When To Trigger

Use this skill when the user asks to:

• run ads or execute advertising campaigns with clear operational next steps
• grow revenue or profit, improve roas, reduce cpa, or optimize budget and bidding
• analyze market, traffic, conversion funnel, and campaign performance signals
• apply this specific capability: policy screening, violation detection, compliant rewrite

Typical trigger keywords:

• ads, advertising, campaign, growth, strategy
• revenue, profit, roi, roas, cpa
• budget, bidding, traffic, conversion, funnel
• meta, googleads, tiktokads, youtubeads, amazonads, shopifyads, dsp

Input Contract

Required:

• business_goal: primary objective (sales, leads, traffic, awareness, retention)
• scope: campaign range, market, timeline, and platform scope
• context: URL, account context, historical performance, or request text

Optional:

• kpi_targets: target cpa, roas, revenue, roi, ltv, cvr
• constraints: budget, policy, brand rules, timeline, resource limits
• platform_preference: preferred channels and priority
• baseline_metrics: existing benchmark metrics

Output Contract

Return an execution-ready result with:

1. Intent Summary (goal, KPI, scope)
2. Findings (key observations and assumptions)
3. Action Plan (prioritized next steps)
4. Risks and Guardrails (what can break and what to monitor)
5. Handoff Payload (structured fields for downstream skills)

Workflow

1. Normalize request and confirm objective.
2. Validate available inputs and list missing critical data.
3. Analyze according to this skill focus: policy screening, violation detection, compliant rewrite.
4. Generate prioritized actions tied to KPI impact.
5. Add platform-specific notes and constraints.
6. Emit a compact handoff payload for execution.

Decision Rules

• If KPI is missing, infer likely primary KPI from goal and mark assumption explicitly.
• If data quality is low, return conservative recommendations and required follow-up checks.
• If platform context is unclear, provide platform-agnostic baseline plus channel variants.
• If policy or account risk appears high, require compliance or account checks before scale.
• If urgency is high and uncertainty is high, prioritize reversible low-risk actions first.

Platform Notes

Primary platform scope:

• Meta (Facebook/Instagram), Google Ads, TikTok Ads, YouTube Ads, Amazon Ads, DSP/programmatic

Guidance:

• Use platform-specific recommendations only when evidence supports them.
• Keep naming explicit: Meta, Google Ads, TikTok Ads, YouTube Ads, Amazon Ads, Shopify Ads, DSP.
• If request is cross-channel, provide channel order and budget split rationale.

Constraints And Guardrails

• Do not fabricate data, performance outcomes, or policy approvals.
• Separate facts from assumptions in every recommendation.
• Keep recommendations measurable and tied to explicit KPIs.
• Avoid irreversible changes without validation checkpoints.

Failure Handling And Escalation

• If required inputs are missing, request concise follow-up fields before final recommendation.
• If data sources conflict, report conflict and provide a safe default path.
• If request implies unsupported account actions, escalate with an exact handoff checklist.
• If compliance risk is detected, route to Ads Compliance Review before launch.

Examples

Example 1: Meta ecommerce optimization

Input:

• Goal: sales growth with lower cpa
• Platform: Meta (Facebook/Instagram)

Output focus:

• top blockers
• prioritized fixes
• week-1 actions and expected KPI movement

Example 2: Google Ads lead generation

Input:

• Goal: improve lead quality and stabilize cpl
• Platform: Google Ads

Output focus:

• search intent structure
• budget and bidding adjustments
• lead-routing handoff fields

Example 3: TikTok plus YouTube scale test

Input:

• Goal: scale traffic while protecting roas
• Platforms: TikTok Ads and YouTube Ads

Output focus:

• test matrix
• risk guardrails
• monitoring and rollback triggers

Quality Checklist

• All required sections are present
• At least 3 registry keywords appear in When To Trigger
• Input and output contracts are explicit and actionable
• Workflow is step-based and execution ready
• Platform references are concrete when applicable
• At least 3 examples are included