AgentSkillsCN

emdb-analytics

查看EmergentDB的分析与使用统计信息。适用于用户希望查看API使用量、延迟、错误、增长趋势,或按键查看详细统计信息的场景。

SKILL.md
--- frontmatter
name: emdb-analytics
description: View EmergentDB analytics and usage stats. Use when the user wants to check API usage, latency, errors, growth, or per-key stats.
allowed-tools: Bash, Read, Write, Edit

EmergentDB Analytics

Help the user retrieve analytics and usage data from their EmergentDB account.

TypeScript SDK

typescript
import { EmergentDB } from "emergentdb";

const db = new EmergentDB("emdb_your_api_key");

// Request stats by endpoint (last 30 days)
const endpoints = await db.analyticsEndpoints();
// [{ endpoint, requestCount, totalBytes, avgLatencyMs, p95LatencyMs, errorCount }]

// Usage by namespace (last 30 days)
const namespaces = await db.analyticsNamespaces();
// [{ namespace, requestCount, totalVectors, avgLatencyMs }]

// Latency percentiles by day (last 30 days)
const latency = await db.analyticsLatency();
// [{ date, p50, p95, p99, requestCount }]

// Error rates by day (last 30 days)
const errors = await db.analyticsErrors();
// [{ date, totalRequests, errorCount, error4xx, error5xx }]

// Per-API-key usage (last 30 days)
const keys = await db.analyticsKeys();
// [{ apiKeyId, keyName, keyPrefix, requestCount, totalBytes, avgLatencyMs, lastUsed }]

// Vector count growth (daily snapshots, last 90 days)
const growth = await db.analyticsGrowth();
// [{ date, vectorCount }]

Python SDK

python
from emergentdb import EmergentDB

db = EmergentDB("emdb_your_api_key")

# Request stats by endpoint (last 30 days)
endpoints = db.analytics_endpoints()
# [EndpointStats(endpoint, requestCount, totalBytes, avgLatencyMs, p95LatencyMs, errorCount)]

# Usage by namespace (last 30 days)
namespaces = db.analytics_namespaces()
# [NamespaceStats(namespace, requestCount, totalVectors, avgLatencyMs)]

# Latency percentiles by day (last 30 days)
latency = db.analytics_latency()
# [LatencyEntry(date, p50, p95, p99, requestCount)]

# Error rates by day (last 30 days)
errors = db.analytics_errors()
# [ErrorEntry(date, totalRequests, errorCount, error4xx, error5xx)]

# Per-API-key usage (last 30 days)
keys = db.analytics_keys()
# [KeyStats(apiKeyId, keyName, keyPrefix, requestCount, totalBytes, avgLatencyMs, lastUsed)]

# Vector count growth (daily snapshots, last 90 days)
growth = db.analytics_growth()
# [GrowthEntry(date, vectorCount)]

Available Analytics

MethodDataWindow
analyticsEndpoints / analytics_endpointsRequests, bytes, latency per endpoint30 days
analyticsNamespaces / analytics_namespacesRequests, vectors, latency per namespace30 days
analyticsLatency / analytics_latencyp50, p95, p99 latency by day30 days
analyticsErrors / analytics_errorsError counts (4xx, 5xx) by day30 days
analyticsKeys / analytics_keysUsage stats per API key30 days
analyticsGrowth / analytics_growthTotal vector count by day90 days

When helping the user, suggest the right analytics method based on what they want to understand (performance, errors, growth, etc.).