AtlasBurn
DocsPlatformIntegrationsOperationsReference

MCP Server

Read-only · OAuth 2.1 (PKCE) · in the official MCP registry

AtlasBurn ships a public Model Context Protocol server so any AI assistant — Claude, ChatGPT, Cursor, Codex — can answer questions from your real AtlasBurn data. It is read-only, OAuth-secured, and listed in the official registry as com.atlasburn/atlasburn. Endpoint: https://app.atlasburn.com/api/mcp.

What it does

Most AI cost tools make you open a dashboard. The MCP server lets you stay in your editor or chat and just ask: "what's my AI runway?" or "have there been any guardrail breaches this week?" — and get answers pulled live from your telemetry. Everything is read-only and content-blind: it never exposes prompts or model outputs, only aggregated economics and metadata.

Tools

ToolReturns
get_cost_summarySpend + token usage over a window (default 30d)
get_cost_breakdownAttribution by model and by feature
get_runway_forecastRunway days, burn, reserves, systemic risk
get_guardrail_statusEnabled/mode/status, budgets, recent breaches
get_runtime_signalsRetry rate, loop probability, RPS, model mix

All five tools are annotated readOnlyHint: true — no tool can change settings, budgets, keys, or delete data.

Connect it

The connection flow is the same across clients: point them at the endpoint, and they discover OAuth automatically and open the AtlasBurn consent screen.

Claude (Desktop / claude.ai) & ChatGPT

Add a custom / remote connector with the URL below. The client runs the OAuth flow, you sign in to AtlasBurn and approve read-only access, and the tools appear in-chat.

HTTPS
http
https://app.atlasburn.com/api/mcp

MCP Inspector (to test)

bash
npx @modelcontextprotocol/inspector
# Transport: Streamable HTTP
# URL: https://app.atlasburn.com/api/mcp
# Connect -> sign in -> Authorize -> List Tools

Authentication & security

  • OAuth 2.1 authorization-code flow with PKCE (S256 only); scoped read-only token (atlasburn:read), short-lived access + rotating refresh.
  • Tenant isolation — every call is scoped to the org resolved from the token, never from a tool argument.
  • Content-blind — only aggregated economics/metadata is exposed; never prompts or completions.
  • Zero writes — the endpoint reads only; codes/tokens are stored as salted HMAC hashes; every tool call is audit-logged.

Behind Cloudflare?

If your app domain sits behind Cloudflare, make sure Bot Fight Mode / "Block AI bots" is not challenging the connector paths — AI platforms call your server from datacenter IPs, and a bot rule will 403 the OAuth registration step.

Registry

AtlasBurn is published in the official MCP registry. Marketplaces that sync from it (Glama, mcp.so, PulseMCP) list it automatically.

bash
curl "https://registry.modelcontextprotocol.io/v0.1/servers?search=com.atlasburn/atlasburn"

Next Steps