Getting it into your agent
It runs from inside its repository, so the clone comes first — what it calls does not travel with the file alone.
git clone --depth 1 https://github.com/krusemediallc/cursor-ad-agentnpx agentmods add skills/krusemediallc/cursor-ad-agent/ad-agent-orchestratorWrote this? Show the measurements
A badge with what this costs and how it scanned, read live from this page, so it follows the numbers instead of freezing them. Markdown for a README, HTML for a documentation site or a project page.
[](https://agentmods.dev/skills/krusemediallc/cursor-ad-agent/ad-agent-orchestrator)<a href="https://agentmods.dev/skills/krusemediallc/cursor-ad-agent/ad-agent-orchestrator"><img src="https://agentmods.dev/badge/skills/krusemediallc/cursor-ad-agent/ad-agent-orchestrator/github.svg" alt="Measured on agentmods" height="20"></a>Or the 80×15 button, for a site that already has a row of RSS and ATOM ones. Only the verdict fits; the numbers stay here.
<a href="https://agentmods.dev/skills/krusemediallc/cursor-ad-agent/ad-agent-orchestrator"><img src="https://agentmods.dev/badge/skills/krusemediallc/cursor-ad-agent/ad-agent-orchestrator.svg" alt="Reviewed on agentmods" width="80" height="20"></a>What it costs to keep this loaded
Counted locally with the o200k_base tokenizer, which is exact for GPT models; Claude uses its own tokenizer and its counts differ. Treat this as one consistent yardstick across the catalogue rather than a bill. Prices are per million input tokens.
| Model | Per session | Once invoked |
|---|---|---|
| Fable 5.1 | $0.00087 | $0.03272 |
| Opus 5 | $0.00044 | $0.01636 |
| Sonnet 5 | $0.00017 | $0.00654 |
| Haiku 4.5 | $0.00009 | $0.00327 |
Grade A, and why
ad-agent-orchestrator scanned grade A with 0 findings against 26 rules in 11 categories — prompt injection, anti-refusal, data exfiltration, privilege escalation, supply chain, agent snooping, system-prompt leakage, SSRF and excessive agency — measured 10d ago.
A static scan of the body, not an audit. Every finding is printed with the line that produced it so you can judge whether it matters here. A mod is markdown that instructs an agent; that is exactly why what it instructs is worth reading.
Nothing flagged
None of the 26 patterns this scan looks for appear in this file: no shell pipes, no recursive deletes, no credential paths, no hidden text, no instruction-override or anti-refusal phrasing, no agent-config snooping. That is not a guarantee, it is the absence of the things that are checkable.
How it starts
The opening of the file, as written. The whole thing — 272 lines — stays where its author put it; the contents beside it link to each section on GitHub.
Ad agent orchestrator
This is the canonical top-level workflow. It owns sequencing, approval gates, lineage, and recovery; component skills own their phase implementations.
Read run-contract.md before opening a run and test-plan.md before proposing concepts or thresholds.
Non-negotiable rules
- A human approves every gate for its exact hashed scope. Approval at one gate never implies approval at another. Any scoped change invalidates approval.
- Never create spend-capable Meta entities as ACTIVE. Deploy ads PAUSED.
Activation is separate, optional, and requires the exact confirmation
ACTIVATEafter reviewing the account, hierarchy, and budgets. - Never run unattended generation, deployment, activation, or delivery mutation. Performance polling is read-only and finite or separately scheduled by the human.
- Regeneration requires approval except for bounded still-image QA-fix retries. After G4, component skills may auto-retry visible image defects with a revised prompt — up to 2 retries per originally requested still (3 attempts total). Disclose possible retry credit exposure at G4 before the first call; each retry is billed. Failed calls, G5 rejection, creative-direction changes, video assets, exhausted QA caps, strategic "try again" requests, and recommendation-driven work need a revised plan, incremental estimate, and fresh applicable gates — never unbounded auto-retry.
- Never invent a claim, receipt, Arcads price, Meta metric, threshold, result, or causal explanation. Unverified copy uses explicit placeholders and cannot be deployed until resolved and approved.
- Meta Ad Library longevity is an adaptation-readiness proxy only. It is not evidence of spend, conversions, ROAS, profitability, or creative performance.
- Keep credentials in environment files. Never put secrets, token-bearing URLs, or authorization headers in prompts, artifacts, commands, or lineage.
- Lineage is append-only. Never edit prior JSONL records or hand-edit a compiled manifest.
What ships with it
5 files beside SKILL.md in the same directory: the scripts, references and assets a skill reads on demand. Not counted in the per-session cost; read them before you install if any of them is executable.
What this file has done since we first saw it
Hashed on every crawl. A supply-chain change to an agent config is a question of when, not whether, so the history is kept rather than the latest state alone.
- 10d ago First seen · 272 lines · 87 tokens per session scan A e1191f157fd2
ad-agent-orchestrator is a skill published in the GitHub repository krusemediallc/cursor-ad-agent (10 stars, last pushed 1mo ago), licensed MIT. It adds 87 tokens to every session and 3,272 once invoked, about $0.0004 per session on Opus 5. A static security scan graded it A with 0 findings. No closer match exists in the catalogue, so it is treated as the original; first seen 2026-08-31.
Other skills, from other repositories
RunAPI Media Generator
A workflow for creating images, videos, music, and audio through RunAPI, a service that provides access to media-generation models. It checks the chosen model and request before submitting the job.
imaginode
Generate AI images and videos through the Imaginode MCP server (imaginode.ai). Use when the user asks to create, generate or illustrate an image, picture, visual, thumbnail, or a short video clip, and the Imaginode MCP tools (generateimage, generatevideo, listmodels, getgenerationstatus) are available.
uni1-image-ad
Use when the user wants to generate a Meta/Facebook image ad with Luma uni-1 and attach it to an existing Meta ad set. Triggers on phrases like "uni-1 ad", "make a Luma ad creative", "new image ad in ", "uni-1 image ad for Meta", "generate uni-1 ad creative", "upload uni-1 image as Meta ad". Anchors on uni-1 + Meta…
image-ad-clone
Use when the user wants to reverse-engineer an existing image ad into a reusable, parameterizable prompt template that can be reused with any brand and any product. Triggers on phrases like "clone this ad as a template", "reverse engineer this ad", "turn this ad into a prompt", "extract a template from this image"…
distribea-media
Generate or place on-brand media (images, video, music, voice) with Distribea. Use whenever a project needs imagery or audio — a hero/illustration/avatar, a page full of placeholders, a video clip, background music, or a voice-over — instead of shipping stock photos, placeholders, CSS-gradient-only sections, or raw…
image-studio
Studio de geracao de imagens inteligente — roteamento automatico entre ai-studio-image (fotos humanizadas/influencer) e stability-ai (arte/ ilustracao/edicao). Detecta o tipo de imagem solicitada...