Harvey Verify
MCP Server
Outcome Verification + Quality Scoring
x402-paid MCP server for post-transaction outcome verification. Uses Grok as LLM-as-judge to validate whether agent service calls delivered expected results. Tracks service quality scores over time.
// CAPABILITIES
What Harvey Verify Does
Post-transaction verification for the agent economy. Trust but verify - every service call, every outcome, scored and tracked.
Outcome Verification
LLM-as-judge validates expected vs actual results with confidence scoring. Grok analyzes whether the service delivered what was promised.
Service Quality Tracking
Rolling quality scores per service and tool over time. Track performance trends and identify degrading services before they cost you.
Outcome Reporting
Report transaction outcomes for community transparency. Build a shared reputation layer that all agents can reference.
Grok-Powered Analysis
xAI Grok provides nuanced reasoning for verification decisions. Not just pass/fail - detailed explanations of why outcomes match or diverge.
Historical Data
Query verification history and quality trends via Supabase. Full audit trail of every verification decision and quality score.
x402 Authentication
No API keys. Pay per verification with USDC on Solana. x402 protocol handles auth and payment in a single HTTP header.
// DOCUMENTATION
Connect & Use
MCP Connection
Available Tools
| Tool | Description | Price |
|---|---|---|
| verify_outcome | LLM-as-judge outcome verification | $0.01 |
| get_service_quality | Query service quality scores | $0.005 |
| report_outcome | Report a transaction outcome | $0.002 |
| get_health | Server health check | Free |
| list_tools | List available tools | Free |
// BUILT WITH
Tech Stack
Verify your agent transactions.
Trust but verify. Every service call scored, every outcome tracked. Pay per verification with USDC.