# Rhumb — Agent-Native Tool Intelligence
> Canonical public docs: https://rhumb.dev/llms.txt

## What is Rhumb?
Rhumb is an agent gateway: Rhumb Index scores and compares services; Rhumb Resolve routes supported capability calls through governed execution rails with receipts.

## Current launchable scope
- Best current fit: research, extraction, generation, and narrow enrichment
- Not the current promise: general business-agent automation or broad multi-system workflow orchestration

## Primary surfaces
- Website: https://rhumb.dev
- API: https://api.rhumb.dev/v1
- MCP: npx -y --package rhumb-mcp@latest rhumb-mcp
- npm: https://www.npmjs.com/package/rhumb-mcp

## Current coverage
- 1,038 scored services across 50+ domains
- 415 capability definitions
- 92 categories
- 16 callable providers
- 21 MCP tools
- 3 execution rails: governed API key, wallet-prefund, x402 / USDC
- 2 operator-controlled credential modes where supported: BYOK, Agent Vault

## Honest current state
- Discovery breadth is wider than execution breadth
- The index covers 1,038 scored services and 415 capability definitions
- Current governed execution surface is 16 callable providers
- Best current fit remains research, extraction, generation, and narrow enrichment

## Discovery (no auth)
- GET https://api.rhumb.dev/v1/search?q={query} — search services
- GET https://api.rhumb.dev/v1/services/{slug}/score — AN Score breakdown
- GET https://api.rhumb.dev/v1/services/{slug}/failures — known failure modes
- GET https://api.rhumb.dev/v1/capabilities — browse capability registry
- GET https://api.rhumb.dev/v1/capabilities/{id}/resolve — ranked providers
- GET https://api.rhumb.dev/v1/leaderboard/{category} — category rankings
- GET https://api.rhumb.dev/v1/telemetry/provider-health — provider health status
- GET https://api.rhumb.dev/v1/pricing — machine-readable pricing

## Execution (requires a live rail)
- POST https://api.rhumb.dev/v1/capabilities/{id}/execute — execute a capability
- GET https://api.rhumb.dev/v1/capabilities/{id}/execute/estimate — estimate the active execution rail, cost, and health before execution; anonymous direct system-of-record paths also preserve machine-readable execute_readiness handoffs

## Execution rails
1. Governed API key: sign up at https://rhumb.dev/auth/login, then send X-Rhumb-Key on the governed rail
2. Wallet-prefund: add balance first, then execute with X-Rhumb-Key on the governed rail
3. x402 / USDC: no signup, pay per call, send X-Payment header

## Operator-controlled credential modes
- BYOK credentials: pass your own upstream API keys when provider control is the point
- Agent Vault: when supported, let Rhumb inject the provider credential at call time while keeping the raw secret behind Rhumb

## MCP tools (21 total)
Discovery: find_services, get_score, get_alternatives, get_failure_modes, discover_capabilities, resolve_capability
Execution: execute_capability, estimate_capability, credential_ceremony, check_credentials, rhumb_list_recipes, rhumb_get_recipe, rhumb_recipe_execute, get_receipt
Billing: budget, spend, check_balance, get_payment_url, get_ledger
Operations: routing, usage_telemetry

## Key tool semantics
- find_services — Search indexed Services by what you need them to do
- get_score — Get the full AN Score breakdown for a Service: execution quality, access readiness, autonomy level, tier label, and freshness
- resolve_capability — Given a Capability ID, and optionally a credential mode, returns ranked providers with health status, cost per call, auth methods, endpoint patterns, execute guidance, and machine-readable recovery fields like recovery_hint.resolve_url, recovery_hint.credential_modes_url, and, when applicable, recovery_hint.alternate_execute_hint or recovery_hint.setup_handoff, plus typo recovery when the capability ID is wrong
- execute_capability — Call a Capability through Rhumb Resolve
- estimate_capability — Estimate the active execution rail, cost, and health before a Capability call; anonymous direct system-of-record paths also preserve machine-readable execute_readiness handoffs
- get_receipt — Retrieve an execution receipt by ID
- usage_telemetry — Get your execution analytics — calls, latency, errors, costs, and provider health for your Rhumb usage

## Pricing
- Discovery: free, no auth
- Execution rails: governed API key, wallet-prefund, or x402 per-call
- Provider-control modes where supported: BYOK and Agent Vault
- No subscriptions, no seat fees, no minimums
- Live pricing and markup terms: https://rhumb.dev/pricing

## Trust and disputes
- Trust overview: https://rhumb.dev/trust
- Methodology: https://rhumb.dev/methodology
- Current self-assessment: https://rhumb.dev/blog/we-scored-ourselves
- Historical baseline: https://rhumb.dev/blog/self-score
- Provider guide and dispute process: https://rhumb.dev/providers
- Public dispute template: https://github.com/supertrained/rhumb/issues/new?template=score-dispute.md
- Public dispute log: https://github.com/supertrained/rhumb/issues?q=is%3Aissue+%22Score+dispute%3A%22
- Private disputes: mailto:providers@supertrained.ai?subject=Score%20Dispute
- Dispute response target: 5 business days

## Links
- Quickstart: https://rhumb.dev/quickstart
- Pricing: https://rhumb.dev/pricing
- Blog: https://rhumb.dev/blog
- GitHub: https://github.com/supertrained/rhumb
- Public agent capabilities: https://rhumb.dev/.well-known/agent-capabilities.json