moxiespirit/oathscore

> Every API makes promises. OathScore checks the receipts.
**The trust layer for AI agents.** Two products:
1. **/now** — A single endpoint returning the current state of the world for trading agents. Exchange status, volatility, events, regime, data health — one call.
2. **OathScore Ratings** — Independent, continuous verification of data API accuracy, uptime, freshness, and reliability. The credit bureau for data APIs.
For AI Agents
For MCP-Compatible Agents (Claude Code, Cursor, etc.)
Requires: pip install httpx mcp[cli] and clone this repo.
MCP Tools
| Tool | Description |
|------|-------------|
| get_now | Current world state: exchanges, volatility, events, data health |
| get_exchanges | Open/close status for 7 exchanges with next transition times |
| get_volatility | VIX, VIX9D, VIX3M, VVIX, SKEW, term structure |
| get_events | Next event, FOMC/CPI countdowns, week high-impact count |
| get_score | OathScore rating for a specific API (0-100 composite + grade) |
| compare_apis | Side-by-side comparison of two or more data APIs |
| get_alerts | Active degradation alerts for monitored APIs |
| check_health | Service health and data freshness |
What OathScore Monitors
For each rated API:
| Metric | Weight | How Measured |
|--------|--------|-------------|
| Accuracy | 35% | Compare forecasts/claims to actual outcomes daily |
| Uptime | 20% | Synthetic monitoring every 60 seconds |
| Freshness | 15% | Is "real-time" actually real-time? |
| Latency | 15% | P50/P95/P99 from multiple regions |
| Schema stability | 5% | Detect breaking changes |
| Documentation | 5% | OpenAPI spec, llms.txt, examples |
| Trust signals | 5% | Published accuracy data, response signing |
Rated APIs (v1)
| API | Category | Score | Status |
|-----|----------|-------|--------|
| Curistat | Futures volatility | -- | Monitoring |
| Alpha Vantage | Equities/macro | -- | Monitoring |
| Polygon.io | Market data | -- | Monitoring |
| Finnhub | Multi-asset | -- | Monitoring |
| Twelve Data | Market data | -- | Monitoring |
| EODHD | Historical data | -- | Monitoring |
| Financial Modeling Prep | Fundamentals | -- | Monitoring |
Scores populate after 30 days of monitoring data.
Machine-Readable Discovery
- /llms.txt — Agent-readable product description
- /llms-full.txt — Complete endpoint documentation
- /.well-known/ai-plugin.json — ChatGPT plugin manifest
- /openapi.json — OpenAPI 3.0.3 spec
- /docs — Interactive Swagger UI
Pricing
| Tier | /now Calls | Score Queries | Price |
|------|-------------|---------------|-------|
| Free | 10/day | 5/day | $0 |
| Founding (first 50) | 5,000/day | 2,500/day | $9/mo (lifetime) |