Catalog › OKF bundles › Marketing & analytics › Google Ads geo targetsGA Google Ads geo targets Bundle OKF 0.2 · 6 conceitos · ndebuhr/agentic-ads-e2e
Open site Project Open in the app JSON README (API)
# Google Ads geo targets Location targeting for Google Ads campaigns: the IDs Google Ads uses for countries and for US states, cities, postal codes, and other places, and how to target with them. To find IDs, search [the CSV](/geotargets.csv) with a one-line `grep`, from the repository root: ```bash grep -i '"Boulder County,Colorado,' .agents/skills/google-ads-geo-targets/assets/geotargets.csv ``` The CSV is 5.4 MB: always search it, never open or load the whole file. [Geo target constants](/geo-target-constants.md) has more searches. ## Data - [Geo target constants](/geo-target-constants.md): the US part of Google's published list of targetable locations (about 64,000), with its columns, statuses, how often it changes, how to search it, and how to update it. - [geotargets.csv](/geotargets.csv): the list itself (5.4 MB). Search it; don't open it. - [Countries](/countries.md): criterion IDs for every active country target. - [US states](/us-states.md): criterion IDs for the 50 US
README Loading documentation…
# agentic-ads-e2e
Agent skills and infrastructure for running Google Ads campaigns end-to-end, from creative generation to campaign reporting.
- `.agents/skills/` — agent skills for image generation, video editing, and Google Ads (campaign summaries, Performance Max asset groups, Search campaigns and responsive search ads, experiments)
- `ads-agent/` — ADK agent that runs the skills on Gemini through Vertex AI (see its README)
- `google-ads-e2e/` — Terraform to deploy the official Google Ads MCP server on Cloud Run (see its README)
## MCP servers
The skills expect these servers in Antigravity. Register them with `agy mcp add`; the config lives in `~/.gemini/config/mcp_config.json`.
**GEAP (Vertex AI) — image generation and prediction.** These use OAuth. Create an OAuth 2.0 *Web application* client in your Google Cloud project with:
- Authorized JavaScript origin: `https://antigravity.google`
- Authorized redirect URI: `https://antigravity.google/oauth-callback`
Then add the servers and attach the client to each entry in `mcp_config.json` (`agy mcp add` has no OAuth flag):
```bash
agy mcp add geap-generate https://aiplatform.googleapis.com/mcp/generate
agy mcp add geap-predict https://aiplatform.googleapis.com/mcp/predict
```
```json
"geap-generate": {
"serverUrl": "https://aiplatform.googleapis.com/mcp/generate",
"oauth": { "clientId": "<CLIENT_ID>.apps.googleusercontent.com", "clientSecret": "<CLIENT_SECRET>" }
}
```
**GenMedia (Veo, Omni, avtool) — video generation and editing.** Local stdio servers. One script installs the binaries, checks ffmpeg, creates the output bucket, and registers them:
```bash
bash .agents/skills/genmedia-video-editing/scripts/setup.sh [--project PROJECT_ID] [--bucket BUCKET]
```
**Google Ads MCP.** Deployed to Cloud Run by the Terraform in `google-ads-e2e/`; its output prints the exact `agy mcp add google-ads-mcp <url>/mcp` command. See that README.
Verify with `agy mcp list`.
## License
MIT. See [LICENSE](https://github.com/ndebuhr/agentic-ads-e2e/blob/HEAD/LICENSE).
Details Type OKF bundles Topic Marketing & analytics Publisher ndebuhr Source okf_github Category Data Version 0.2 Last update 2026-10-18T03:11:19Z Project status Active Language Python License MIT Added 2026-10-16 15:24:11 Updated 2026-10-16 15:24:11 Source id ndebuhr/agentic-ads-e2e:ads-agent/.agents/skills/google-ads-geo-targets/assets/index.mdMore