Using Agent Safe (or trying it)? Tell us what your agent pays for

[nicolastursi7-ops]

If you've installed any of the packages (agent-pay, the MCP server, or the LangChain, AI SDK or AgentKit tools) and made a payment with it, even one on testnet, we'd like to hear about it here. There's no telemetry in the packages, so this thread is the only way we learn who uses them.

Useful to us, answer only what you like:

• What the agent does and which framework it runs in.
• Which sellers it pays (the x402 APIs or MCP servers). We have guides for 34 of them in docs/integrations, and we'll write one for a seller that's missing.
• What got refused that shouldn't have been, or what passed that shouldn't have. Include the refusal code if you have it.
• What's missing before you'd run it with real money.

What we can do in return:

• Test your seller live and add a verified guide for it.
• Help you set up the registry and the sealed plan for your case.
• List your project in the README as a user, only if you ask for it.

If you'd rather not post in public, write to hello@deepfirstsearch.com.