Compliance Friendly Email Testing for AI Agents
Subject line experiments shouldn't force you to compromise on data hygiene or sender reputation. marginal gives AI agents a controlled way to run compliance friendly email testing — measurable, auditable, and built on managed sends.
What makes email testing compliance friendly
Compliance friendly email testing means your experiments stay within the boundaries that keep deliverability and trust intact: testing on opted-in audiences, tracking results transparently, and avoiding deceptive subject lines that trigger spam complaints.
marginal is a hosted email MCP server, so the send infrastructure and tracking are managed for you. Your agent describes the experiment; marginal handles the launch, the open/click measurement, and the within-test lift math without exposing raw subscriber data to the model.
- Managed sends keep delivery consistent and reputation-aware
- Open and click tracking is recorded per variant, not scraped ad hoc
- Subject line A/B tests compare honest variants, not clickbait swaps
- Results are quantified as lift, so decisions are evidence-based
How an AI agent runs a subject line A/B test
AI agent email marketing works best when the agent reasons about copy and a dedicated tool owns the experiment mechanics. marginal exposes four MCP tools that map cleanly to that workflow.
- generate_variants — draft candidate subject lines from a prompt
- launch_test — start a subject line A/B test on your audience
- get_results — pull open/click and within-test lift metrics
- recommend_next — suggest the next experiment based on outcomes
Keeping experiments auditable
Because every test runs through marginal's endpoint at https://marginal.sh/mcp with a Bearer API key, there's a clear record of what was sent and how it performed. That auditability is the practical core of compliance friendly email testing — you can show what variant went to which segment and what lift it produced.
Authentication uses a single marg_live_ key, and the free tier covers 100 experiments per month, giving teams room to validate a responsible testing process before scaling.
Where it fits in your stack
marginal connects to the MCP clients your team already uses, so the agent driving your subject line A/B test can live wherever you work. Add the email MCP server endpoint and your key, and the testing tools appear alongside your other agent capabilities.
- Cursor, Claude Desktop, Claude Code, and Windsurf
- OpenAI Codex, Cline, Continue, and Zed
- Full setup details in the docs at https://marginal.sh/docs/
Get started with marginal
marginal is a hosted email marketing MCP server at marginal.sh. Sign up free, create an API key, and connect https://marginal.sh/mcp from Cursor, Codex, or Claude Desktop.
- 100 experiments/month on the free tier
- Four MCP tools: generate_variants, launch_test, get_results, recommend_next
- Listed in the MCP Registry as sh.marginal/mcp