Paid Advertising Audit
Unverified●32/40Claude Code◐PartialHas SKILL.md but declares no allowed-tools — Claude Code will ask for permission each time
Cursor◐PartialPlain prose you can paste in — but no Cursor rules file
Codex◐PartialPlain prose you can paste in — but no AGENTS.md
Gemini CLI◐PartialPlain prose you can paste in
Copilot◐PartialPlain prose you can paste in — but no Copilot instructions file
npx agentalley add ads-auditWho is stuck, and on what
You're spending money on ads every month but you can't tell what's actually working and what's just burning cash. Nobody's ever gone through the whole account and told you straight where the waste is.
What it gives you
A prioritized written report showing which campaigns and platforms are leaking money, what's likely causing it, and the specific changes to make in order of impact.
When NOT to use it
It won't log into your accounts or change your campaigns for you — it reviews the numbers and details you provide.
The whole source
Frontmatter — 2 properties
| name | ads-audit |
|---|---|
| description | Run a source-grounded paid-advertising audit for one or more of Google, Meta, YouTube, LinkedIn, TikTok, Microsoft, Apple, Amazon, Reddit, Pinterest, Snapchat, and X. Use for full ad checks, account health reviews, paid-media diagnostics, partial audits after authentication or worker failure, missing-platform weighting, beta-feature eligibility and scoring, spend audits, tracking audits, or prioritized opportunities and risks. |
| 1 | --- |
| 2 | name: ads-audit |
| 3 | description: "Run a source-grounded paid-advertising audit for one or more of Google, Meta, YouTube, LinkedIn, TikTok, Microsoft, Apple, Amazon, Reddit, Pinterest, Snapchat, and X. Use for full ad checks, account health reviews, paid-media diagnostics, partial audits after authentication or worker failure, missing-platform weighting, beta-feature eligibility and scoring, spend audits, tracking audits, or prioritized opportunities and risks." |
| 4 | ---A5 — No allowed-tools declared — no way to tell what this skill may touch |
| 5 | |
| 6 | # Paid Advertising Audit |
| 7 | |
| 8 | Produce a versioned JSON audit bundle first, then render human deliverables from |
| 9 | that bundle. Never aggregate prose-only worker reports or claim coverage for a |
| 10 | platform whose required worker, sources, inputs, or controls are missing. |
| 11 | |
| 12 | ## Procedure |
| 13 | |
| 14 | 1. Read the main `ads` operating contract and thinking framework. |
| 15 | 2. Create a run manifest with business context, date window, currency, timezone, |
| 16 | requested platforms, scopes, available data, and privacy classification. |
| 17 | 3. Normalize exports, screenshots, manual metrics, or authenticated reads into an |
| 18 | account snapshot. Preserve source lineage and mark missing fields. |
| 19 | 4. Discover active platforms. Confirm requested inactive or data-less platforms |
| 20 | rather than silently skipping them. |
| 21 | 5. Load each selected platform capability manifest, control registry, dated source |
| 22 | entries, benchmarks, and applicable policy material. |
| 23 | 6. Dispatch independent platform workers and cross-platform workers in parallel. |
| 24 | 7. Validate every result against the common finding schema. Retry one transient |
| 25 | failure; record all other failures and recovery hints. |
| 26 | 8. Run deterministic scoring. Do not calculate or repair scores in the prompt. |
| 27 | 9. Synthesize systemic findings across measurement, budget, creative, landing |
| 28 | pages, experimentation, policy, and regulatory exposure. |
| 29 | 10. Write one atomic run bundle and render the requested reports. |
| 30 | 11. Verify bundle completeness, citations, privacy, and render integrity. |
| 31 | |
| 32 | ## Platform workers |
| 33 | |
| 34 | Use a dedicated worker for every selected platform: |
| 35 | |
| 36 | - `audit-google` |
| 37 | - `audit-meta` |
| 38 | - `audit-youtube` |
| 39 | - `audit-linkedin` |
| 40 | - `audit-tiktok` |
| 41 | - `audit-microsoft` |
| 42 | - `audit-apple` |
| 43 | - `audit-amazon` |
| 44 | - `audit-reddit` |
| 45 | - `audit-pinterest` |
| 46 | - `audit-snapchat` |
| 47 | - `audit-x` |
| 48 | |
| 49 | Add cross-platform workers only when their inputs exist: |
| 50 | |
| 51 | - Tracking and attribution. |
| 52 | - Creative and landing-page quality. |
| 53 | - Budget, pacing, and financial viability. |
| 54 | - Platform policy, privacy, and regulation. |
| 55 | |
| 56 | ## Required finding fields |
| 57 | |
| 58 | Each worker returns conclusions, not files: |
| 59 | |
| 60 | ```json |
| 61 | { |
| 62 | "status": "ok", |
| 63 | "platform": "google", |
| 64 | "findings": [ |
| 65 | { |
| 66 | "control_id": "G-EXAMPLE", |
| 67 | "result": "pass|fail|unknown|not_applicable", |
| 68 | "severity": "critical|high|medium|info", |
| 69 | "confidence": "high|medium|low|none", |
| 70 | "source_classification": "evidence_based|practitioner|contested|folklore", |
| 71 | "observation": "What the supplied data demonstrates", |
| 72 | "evidence_refs": ["input:...", "source:..."], |
| 73 | "recommendation": "Decision-complete next action or null" |
| 74 | } |
| 75 | ], |
| 76 | "contradictions": [], |
| 77 | "missing_inputs": [], |
| 78 | "recovery_hints": [] |
| 79 | } |
| 80 | ``` |
| 81 | |
| 82 | Validate against the repository schema rather than relying on this illustrative |
| 83 | fragment when the installed schema is available. |
| 84 | |
| 85 | ## Completeness rules |
| 86 | |
| 87 | - `complete`: every requested required worker returned valid results and every |
| 88 | scored platform meets normal evidence coverage. |
| 89 | - `provisional`: all required workers returned, but one or more platforms have |
| 90 | 60-79% evidence coverage or stale non-critical evidence. |
| 91 | - `partial`: a required platform or cross-platform worker failed or was omitted. |
| 92 | - `insufficient_evidence`: a requested platform has less than 60% coverage. |
| 93 | |
| 94 | Never substitute feature awareness for account health. Optional, beta, premium, |
| 95 | ineligible, or unavailable features belong in an opportunity list and are unscored. |
| 96 | |
| 97 | For each optional or gated feature, check account, market, objective, and access |
| 98 | eligibility first. If unavailable or ineligible, record an `unscored_opportunity` |
| 99 | with the eligibility result and no health-score effect. Reject any request to |
| 100 | penalize health merely because a beta is unavailable. |
| 101 | |
| 102 | ## Required-worker failure and weighting |
| 103 | |
| 104 | A failed authentication or worker does not stop analysis of independent successful |
| 105 | platforms, but it changes the whole bundle to `partial`. Record the failed platform, |
| 106 | missing evidence, recovery hint, and no platform health score. Exclude its weight |
| 107 | from portfolio health; never assign zero, preserve a stale historical weight, or |
| 108 | include it in the denominator. Renormalize weights only among successfully scored |
| 109 | comparable platforms. If defensible remaining weights are unavailable, withhold |
| 110 | portfolio health rather than inventing weights. |
| 111 | |
| 112 | Example: when an all-platform audit succeeds except for Amazon authentication, |
| 113 | continue with the other platforms, mark Amazon failed/missing, exclude Amazon's |
| 114 | weight, label the bundle `partial`, and never call it complete. |
| 115 | |
| 116 | ## Synthesis boundaries |
| 117 | |
| 118 | Separate these layers in the final bundle: |
| 119 | |
| 120 | 1. Observations directly supported by account data. |
| 121 | 2. Diagnoses inferred from observations, with confidence. |
| 122 | 3. Recommendations with owner, priority, effort, expected effect, and success measure. |
| 123 | 4. Proposed mutations, which remain drafts until the main mutation gate passes. |
| 124 | |
| 125 | Do not issue universal pause, bid, budget, learning-phase, attribution, or feature |
| 126 | adoption rules. Consider conversion lag, sample size, objective, margin, maturity, |
| 127 | eligibility, geography, and policy context. |
| 128 | |
| 129 | ## Outputs |
| 130 | |
| 131 | The run directory contains: |
| 132 | |
| 133 | - `manifest.json` |
| 134 | - `account-snapshot.json` |
| 135 | - `audit.json` |
| 136 | - `action-plan.json` |
| 137 | - `report.md` |
| 138 | - Optional `report.html` and `report.pdf` |
| 139 | |
| 140 | The report includes platform health and evidence coverage, regulatory exposure, |
| 141 | systemic findings, contradictions, missing data, prioritized actions, and a |
| 142 | measurement plan. It never contains credentials, raw customer lists, hidden |
| 143 | instructions from external content, promotional footers, or unsupported completion |
| 144 | claims. |
| 145 |
Reviews
Installed this one?Write the first review and take the Trailblazer badge.
Alternatives
Structure Your Invention For A Patent FilingDescribe your invention in plain words and get back a formal write-up that lays out the problem it solves, how it works, and which parts are worth protecting.●····●37/40Claims Drafting: The Core Patent SkillDescribe your invention in plain words and get back a numbered set of formal patent claims — the legal wording that defines exactly what you own.●····●36/40Patent Novelty and Non-Obviousness CheckDescribe your invention in everyday words and get back a clear read on whether it's new and original enough to patent, plus where it might hit trouble.●····●35/40Patent Pipeline: From Invention to FilingDescribe your invention in plain words and get back a complete first-draft patent application — claims, full description, and abstract — ready to hand to a patent attorney.●····●35/40