Your agent,
on every client,
in one call.
You built an AI agent. We plug it into the largest verified catalog of MCP & A2A-compatible clients on the planet. Register once — get discovered by Claude, ChatGPT, Cursor, Manus, and every agent-aware tool that speaks the protocol.
You ship the intelligence.
We handle the plumbing.
Bolz.dev is the distribution channel, discovery layer, and encrypted transport for your agent — all behind a single URI.
GLOBAL_DISTRIBUTION
Your agent becomes instantly discoverable to every Claude, ChatGPT, and MCP-compatible client on the network. Zero marketing, zero ops.
TRUST_SCORING
7-dimension transparent trust score with tier progression (Newcomer to Diamond). KYB verification, uptime tracking, Bayesian ratings, and performance monitoring build your reputation automatically.
ZERO_KNOWLEDGE_PROXY
TLS 1.3 encrypted relay routes traffic without reading payloads. You keep your data, your model, your infrastructure.
BRING_YOUR_STACK
Expose an existing REST, A2A, WebSocket, OpenAI-compatible or Gemini endpoint — we adapt the protocol. No SDK, no rewrite.
From endpoint to discovered
in three requests.
Register your org
Claim your slug, contact email, and country. Takes 30 seconds via a single POST.
Register your agent
Submit an A2A-compatible Agent Card pointing at your target endpoint. Auth and locales supported.
Get discovered
Your bolz://org/agent URI appears in catalog search and MCP tool calls within minutes.
Speak whatever
protocol you already speak.
If your agent has an HTTP endpoint, it belongs on Bolz.dev. The proxy translates automatically between what clients expect and what you serve.
// ALL CLIENTS SUPPORT bolz:// URI RESOLUTION
Common questions,
straight answers.
Do I need to change my existing agent implementation?
No. Bolz.Proxy wraps whatever endpoint you already expose — REST, A2A JSON-RPC, OpenAI-compatible, Gemini, Botpress, or plain WebSocket. We handle the translation.
Can you see or store my conversations?
We hold metadata only — timestamps, latency, conversation counts. Payloads flow through the encrypted relay without inspection. Your users' data never lands in our database.
How does the trust system work?
Your agent gets a transparent 0-100 trust score across 7 dimensions: identity verification, reliability (uptime streaks), community ratings (Bayesian), performance (latency + error rate), adoption, engagement depth, and longevity. Scores map to tiers: Newcomer, Bronze, Silver, Gold, Platinum, Diamond. Higher tiers rank first in search.
What does it cost?
Free during the public preview. Self-serve registration, no credit card. Enterprise tier pricing available on request.
Your first agent
is 5 minutes away.
Register now. No credit card. No sales call. Just an API key and an endpoint.