flin-meta-ads-mcp
Read-only MCP server for Meta Ads that lists and reads ad accounts, campaigns, ad sets, ads, ad images, creatives, and fetches insights at various levels.
README
flin-meta-ads-mcp
flin-meta-ads-mcp is a public, strict read-only MCP server for Meta Ads. It is designed to be installed with uvx and loaded into Claude with minimal setup.
What it does
- Lists and reads Meta Ads accounts
- Reads campaigns, ad sets, ads, ad images, and creatives
- Gets ad previews (including
preview_urllinks) - Fetches insights for account, campaign, ad set, and ad levels
- Stays read-only in
v0.1.x
Scope for v0.1.x
This release is strictly read-only.
That means:
- No create, update, pause, resume, or delete actions
- No OAuth flow or token refresh
- No generic Graph API proxy
Install
Install and run the latest published PyPI release:
uvx flin-meta-ads-mcp
Install and refresh to the newest published release:
uvx --refresh flin-meta-ads-mcp
Run the latest GitHub code (main branch):
uvx --refresh --from git+https://github.com/flin-agency/flin-meta-ads-mcp.git flin-meta-ads-mcp
Claude config
Stable (latest published PyPI release, refreshed):
{
"mcpServers": {
"flin-meta-ads-mcp": {
"command": "uvx",
"args": ["--refresh", "flin-meta-ads-mcp"],
"env": {
"META_ACCESS_TOKEN": "EAA...",
"META_GRAPH_API_VERSION": "v21.0"
}
}
}
}
Latest GitHub code (main branch, refreshed):
{
"mcpServers": {
"flin-meta-ads-mcp": {
"command": "uvx",
"args": [
"--refresh",
"--from",
"git+https://github.com/flin-agency/flin-meta-ads-mcp.git",
"flin-meta-ads-mcp"
],
"env": {
"META_ACCESS_TOKEN": "EAA...",
"META_GRAPH_API_VERSION": "v21.0"
}
}
}
}
Environment variables
Required:
META_ACCESS_TOKEN: Meta access token with read permissions
Optional:
META_GRAPH_API_VERSION: Graph API version, defaultv21.0META_TIMEOUT_SECONDS: request timeout, default30META_MAX_RETRIES: retry count for transient failures, default3META_DEFAULT_AD_ACCOUNT_ID: optional fallback ad account (e.g.act_123...) used when a tool call omitsad_account_idRUN_LIVE_META_TESTS: set to1to enable live integration tests
Ad account selection:
- If the token has exactly one accessible ad account, the server resolves it automatically.
- If
META_DEFAULT_AD_ACCOUNT_IDis set, the server uses it whenad_account_idis omitted. - If the token has multiple accessible ad accounts, pass
ad_account_idin the tool call.
2-minute smoke test
Use these first three tool calls to confirm the server is working:
list_ad_accountslist_campaignsget_insightswithlevel=campaignanddate_preset=last_7d
Example flow in Claude:
Call list_ad_accounts
Call list_campaigns
Call get_insights with level=campaign and date_preset=last_7d
If the first call works but later calls fail, the issue is usually permissions or ad account scope.
If list_campaigns returns a selection request, call it again with one of the suggested ad_account_id values.
Troubleshooting
| Problem | Likely cause | Fix |
|---|---|---|
| Token missing | META_ACCESS_TOKEN is not set |
Add the env var and restart Claude |
| Token invalid | Expired or wrong token | Generate a valid Meta read token |
| Permission denied | Missing ads_read or account access |
Grant the token access to the ad account |
| Ambiguous ad account | Token can access multiple ad accounts | Pass ad_account_id per tool call |
| Rate limit errors | Meta API throttling | Retry later or reduce the number of insight calls |
Development
Local development quickstart:
python -m pip install -e ".[dev]"
pytest -q
ruff check .
mypy src
If you want to run live Meta API tests:
RUN_LIVE_META_TESTS=1 pytest -q
Release
Create and push a release tag:
git tag v0.1.9
git push origin v0.1.9
After release, users can run:
uvx flin-meta-ads-mcp
Notes
- Use only ad accounts you are allowed to access
- This server is intended for analysis and reporting, not mutation
- Public users should treat
META_ACCESS_TOKENas a secret and supply it through environment variables only
Recommended Servers
playwright-mcp
A Model Context Protocol server that enables LLMs to interact with web pages through structured accessibility snapshots without requiring vision models or screenshots.
Magic Component Platform (MCP)
An AI-powered tool that generates modern UI components from natural language descriptions, integrating with popular IDEs to streamline UI development workflow.
Audiense Insights MCP Server
Enables interaction with Audiense Insights accounts via the Model Context Protocol, facilitating the extraction and analysis of marketing insights and audience data including demographics, behavior, and influencer engagement.
VeyraX MCP
Single MCP tool to connect all your favorite tools: Gmail, Calendar and 40 more.
graphlit-mcp-server
The Model Context Protocol (MCP) Server enables integration between MCP clients and the Graphlit service. Ingest anything from Slack to Gmail to podcast feeds, in addition to web crawling, into a Graphlit project - and then retrieve relevant contents from the MCP client.
Kagi MCP Server
An MCP server that integrates Kagi search capabilities with Claude AI, enabling Claude to perform real-time web searches when answering questions that require up-to-date information.
E2B
Using MCP to run code via e2b.
Neon Database
MCP server for interacting with Neon Management API and databases
Exa Search
A Model Context Protocol (MCP) server lets AI assistants like Claude use the Exa AI Search API for web searches. This setup allows AI models to get real-time web information in a safe and controlled way.
Qdrant Server
This repository is an example of how to create a MCP server for Qdrant, a vector search engine.