Python SDK for AI agent monitoring, LLM cost tracking, benchmarking, and more. Integrates with most LLMs and agent frameworks including CrewAI, Agno, OpenAI Agents SDK, Langchain, Autogen, AG2, and CamelAI
Python agentagentopsagents-sdkaianthropicautogencost-estimationcrewaievalsevaluation-metricsgroqlangchainllmmistralollamaopenaiopenai-agents
import { Tables } from '@/lib/types_db';
export interface ISessionAndStats extends Omit<Tables<'sessions'>, 'host_env'> {
host_env: HostEnv | null;
stats: Tables<'stats'>;
}
export type HostEnv = {
OS?: OS;
RAM?: RAM;
CPU?: CPU;
SDK?: SDK;
Disk?: Disk;
'Installed Packages'?: InstalledPackages;
'Virtual Environment'?: VirtualEnvironment;
'Project Working Directory'?: ProjectWorkingDirectory;
};
type OS = {
OS: string;
Hostname: string;
'OS Release': string;
'OS Version': string;
};
type RAM = {
Used: string;
Total: string;
Available: string;
Percentage: string;
};
type CPU = {
'CPU Usage': string;
'Total cores': number;
'Physical cores': number;
};
type SDK = {
'Python Version': string;
'System Packages': SystemPackages;
'AgentOps SDK Version': string;
};
export type SystemPackages = {
[key: string]: string;
};
type Disk = {
[key: string]: {
Free: string;
Used: string;
Total: string;
Mountpoint: string;
Percentage: string;
};
};
type InstalledPackages = {
'Installed Packages': {
[key: string]: string;
};
};
type VirtualEnvironment = {
'Virtual Environment': string;
};
type ProjectWorkingDirectory = {
'Project Working Directory': string;
};
/*
* Prompt Messages and Completion Messages are validated against a json schema before
* being stored in Supabase. While Supabase only knows that the fields are JSON, we
* know their full shapes.
*/
export type llms = Omit<Tables<'llms'>, 'prompt' | 'completion'> & {
prompt: prompt;
completion: completion;
};
export type prompt = chatmlPrompt | promptCompletionString;
export type completion = {
type: 'string' | 'chatml';
string?: string;
messages?: chatmlTextMessage;
};
export type chatmlPrompt = {
type: 'chatml';
messages: chatmlTextMessage[] | chatmlImageMessage[];
};
export type toolCall = {
id: string;
type: 'function';
function: {
name: string;
arguments: string;
};
};
export type functionCall = toolCall;
type chatmlTextMessage = {
role: string;
content: string;
tool_calls?: {
id: string;
function: {
name: string;
arguments: string;
};
}[];
};
type chatmlImageMessage = {
role: 'user' | 'assistant';
content:
| {
type: 'text';
text: string;
}
| {
type: 'image_url';
image_url: {
url: string;
};
};
};
type promptCompletionString = {
type: 'string';
string: string;
};
export interface IExtendedSession {
id: string;
agents: Tables<'agents'>[] | null;
threads: Tables<'threads'>[] | null;
llms: llms[] | null;
actions: Tables<'actions'>[] | null;
tools: Tables<'tools'>[] | null;
errors: Tables<'errors'>[] | null;
}
export interface ILlms extends llms {
type: 'LLM Call';
}
export interface ITools extends Tables<'tools'> {
type: 'Tool';
}
export interface IActions extends Tables<'actions'> {
type: 'Action';
}
export interface IErrors extends Tables<'errors'> {
type: 'Error';
}
export type IEvent = IActions | ILlms | ITools;
/**
* Pagination parameters for API requests
*/
export interface PaginationParams {
limit?: number;
offset?: number;
}
/**
* Session representation
*/
export interface Session {
id: string;
projectId: string;
startTime: Date;
endTime: Date | null;
durationMs: number;
endState: string;
tags: string[];
metadata: Record<string, any>;
}
/**
* Trace based on OpenTelemetry span
*/
export interface Trace {
traceId: string;
spanId: string;
parentSpanId: string | null;
name: string;
startTime: Date;
endTime: Date;
durationNs: number;
statusCode: 'UNSET' | 'OK' | 'ERROR';
statusMessage: string;
attributes: Record<string, any>;
resourceAttributes: Record<string, any>;
events: TraceEvent[];
projectId: string;
}
/**
* Event within a trace
*/
export interface TraceEvent {
name: string;
timestamp: Date;
attributes: Record<string, any>;
}
/**
* Session log based on OpenTelemetry log record
*/
export interface SessionLog {
timestamp: Date;
traceId: string | null;
spanId: string | null;
severityText: string;
severityNumber: number;
body: string | Record<string, any>;
attributes: Record<string, any>;
resourceAttributes: Record<string, any>;
}
/**
* Session metric based on OpenTelemetry metric
*/
export interface SessionMetric {
timestamp: Date;
name: string;
description: string;
unit: string;
dataType: 'GAUGE' | 'SUM' | 'HISTOGRAM' | 'EXPONENTIAL_HISTOGRAM' | 'SUMMARY';
value: number | Record<string, any>;
attributes: Record<string, any>;
resourceAttributes: Record<string, any>;
}
/**
* Session statistics
*/
export interface SessionStats {
sessionId: string;
projectId: string;
startTime: Date;
endTime: Date | null;
durationMs: number;
endState: string;
traceCount: number;
logCount: number;
metricCount: number;
errorCount: number;
}
/**
* OpenTelemetry severity numbers
* Based on https://opentelemetry.io/docs/specs/otel/logs/data-model/#field-severitynumber
*/
export enum SeverityNumber {
UNSPECIFIED = 0,
TRACE = 1,
TRACE2 = 2,
TRACE3 = 3,
TRACE4 = 4,
DEBUG = 5,
DEBUG2 = 6,
DEBUG3 = 7,
DEBUG4 = 8,
INFO = 9,
INFO2 = 10,
INFO3 = 11,
INFO4 = 12,
WARN = 13,
WARN2 = 14,
WARN3 = 15,
WARN4 = 16,
ERROR = 17,
ERROR2 = 18,
ERROR3 = 19,
ERROR4 = 20,
FATAL = 21,
FATAL2 = 22,
FATAL3 = 23,
FATAL4 = 24,
}
/**
* OpenTelemetry severity text values
* Based on https://opentelemetry.io/docs/specs/otel/logs/data-model/#field-severitytext
*/
export enum SeverityText {
TRACE = 'TRACE',
DEBUG = 'DEBUG',
INFO = 'INFO',
WARN = 'WARN',
ERROR = 'ERROR',
FATAL = 'FATAL',
}
/**
* OpenTelemetry span status codes
* Based on https://opentelemetry.io/docs/specs/otel/trace/api/#set-status
*/
export enum SpanStatusCode {
UNSET = 'UNSET',
OK = 'OK',
ERROR = 'ERROR',
}
/**
* OpenTelemetry metric data types
* Based on https://opentelemetry.io/docs/specs/otel/metrics/data-model/
*/
export enum MetricDataType {
GAUGE = 'GAUGE',
SUM = 'SUM',
HISTOGRAM = 'HISTOGRAM',
EXPONENTIAL_HISTOGRAM = 'EXPONENTIAL_HISTOGRAM',
SUMMARY = 'SUMMARY',
}
export interface ProjectMetrics {
project_id: string;
trace_count: number;
span_count: {
total: number;
success: number;
fail: number;
unknown: number;
indeterminate: number;
};
token_metrics: {
prompt_tokens: number;
completion_tokens: number;
total_tokens: {
all: number;
success: number;
fail: number;
};
total_cost: number | string;
average_cost_per_session: string;
avg_tokens: {
all: string;
success: string;
fail: string;
};
by_model?: Record<string, number>;
by_system?: {
llm: number;
[key: string]: number;
};
};
duration_metrics: {
min_duration_ns: number;
max_duration_ns: number;
avg_duration_ns: number;
total_duration_ns: number;
};
success_datetime?: string[];
fail_datetime?: string[];
indeterminate_datetime?: string[];
spans_per_trace?: Record<string, number>[];
failed_traces_dates?: Record<string, number>[];
trace_cost_dates?: Record<string, number>[];
trace_durations?: number[];
freeplan_truncated?: boolean;
}
export interface TraceMetrics {
trace_id: string;
token_metrics: {
prompt_tokens: number;
completion_tokens: number;
total_tokens: number;
total_cost: number;
by_model: Record<string, number>;
by_system: {
llm: number;
[key: string]: number;
};
};
start_time: string;
end_time: string;
total_duration_ns: number;
}