MCPs › Data & APIs › TrustHarness
Enables deterministic security testing of AI agents that use tools by serving synthetic MCP environments with poisoned data, fake secrets, and privileged actions. Records agent tool calls and evaluates security invariants (e.g., canary leaks, forbidden access, approval binding) without an LLM judge or real systems.
Not monetized yet
Turn TrustHarness’s tool calls into revenue: one disclosed sponsored slot, 70% revenue share, fail-open by design.
Install TrustHarness
For anyone using TrustHarness — no Lulu account needed# stdio server — install per the repository README: https://github.com/ofirtro/trustharness
9 field-tested tactics as a designed playbook plus skills your coding agent can run. Free.
Get the Kit →FAQ
TrustHarness installs from source — follow the repository README.
Unrated out of 100, computed from cross-registry traction signals (installs, stars, registry presence) — never influenced by sponsorship.
Similar servers
Works well together
Your server?
This is for the person who owns TrustHarness — adds Lulu Ads to your own code. Not the install steps above, those are for your users.
Copies a ready prompt: your coding agent installs the lulu-ads SDK, wires the slot, and applies the widget design guide.
or set up manually at getlulu.dev/publishers
Get verified so you can edit the page. Your badge is already live below — no claim needed for that.
Managed hosting with monetization built in — waitlist.
[](https://getlulu.dev/mcps/trustharness)