# MeshAI Documentation ## Docs - [Introduction](https://docs.meshai.dev/getting-started/introduction.md): MeshAI is the Agent Control Plane: see, govern, and comply - [Quick Start](https://docs.meshai.dev/getting-started/quickstart.md): Get monitoring in under 5 minutes - [MCP Server](https://docs.meshai.dev/getting-started/mcp-server.md): Governance and observability for Claude Desktop, VS Code, JetBrains, Cursor, Windsurf, Zed, Warp, and 10+ more tools - [Claude Code Connector](https://docs.meshai.dev/getting-started/claude-code.md): Durable, evidence-grade telemetry for Claude Code, built for EU AI Act Article 12 - [Codex CLI Connector](https://docs.meshai.dev/getting-started/codex.md): Zero-code telemetry from OpenAI Codex CLI's built-in OpenTelemetry exporter - [OpenTelemetry Ingestion](https://docs.meshai.dev/getting-started/opentelemetry.md): Send traces, metrics, and logs from any OTLP-compatible agent to MeshAI - [Core Concepts](https://docs.meshai.dev/getting-started/concepts.md): Key concepts in the MeshAI Agent Control Plane - [FAQ](https://docs.meshai.dev/getting-started/faq.md): Frequently asked questions about MeshAI - [Proxy Overview](https://docs.meshai.dev/proxy/overview.md): Zero-code monitoring: one env var, full observability - [Proxy Setup](https://docs.meshai.dev/proxy/setup.md): Configure the MeshAI proxy for your agents - [Provider Configuration](https://docs.meshai.dev/proxy/providers.md): Setup guides for each LLM provider - [OpenAI](https://docs.meshai.dev/framework-guides/openai.md): Auto-track all OpenAI API calls - [Anthropic](https://docs.meshai.dev/framework-guides/anthropic.md): Auto-track all Anthropic Claude API calls - [CrewAI](https://docs.meshai.dev/framework-guides/crewai.md): Auto-track all LLM calls in CrewAI crews - [LangChain / LangGraph](https://docs.meshai.dev/framework-guides/langchain.md): Auto-track LLM calls in LangChain chains and LangGraph agents - [AutoGen](https://docs.meshai.dev/framework-guides/autogen.md): Auto-track LLM calls in AutoGen multi-agent conversations - [Google Gemini](https://docs.meshai.dev/framework-guides/gemini.md): Auto-track all Google Gemini API calls - [AWS Bedrock](https://docs.meshai.dev/framework-guides/bedrock.md): Auto-track all AWS Bedrock LLM calls - [LlamaIndex](https://docs.meshai.dev/framework-guides/llamaindex.md): Auto-track LLM calls in LlamaIndex pipelines - [Agno](https://docs.meshai.dev/framework-guides/agno.md): Auto-track LLM calls in Agno (formerly Phidata) agents - [Pydantic AI](https://docs.meshai.dev/framework-guides/pydantic-ai.md): Auto-track LLM calls in Pydantic AI agents - [Semantic Kernel](https://docs.meshai.dev/framework-guides/semantic-kernel.md): Auto-track LLM calls in Microsoft Semantic Kernel - [Governance Overview](https://docs.meshai.dev/governance/overview.md): Unified governance for AI agents - policies, approvals, audit trail, risk classification, and compliance - [Policies](https://docs.meshai.dev/governance/policies.md): Create and manage governance policies evaluated against agent activity across the proxy and OTLP ingest - [Sanitized Input Capture](https://docs.meshai.dev/governance/payload-capture.md): Opt in to bounded, redacted input evidence from OTLP spans and proxy requests - [Prompt Injection Detection](https://docs.meshai.dev/governance/prompt-injection.md): Always-on scanning at the proxy that blocks prompt injection attacks before they reach LLM providers - [PII Filter](https://docs.meshai.dev/governance/pii-filter.md): Redact, block, or observe personal data in LLM responses at the proxy, with detections recorded as evidence - [Agent Kill Switch](https://docs.meshai.dev/governance/kill-switch.md): Instantly block any agent from making LLM requests - enforced in real-time at the proxy layer - [Shadow Agent Quarantine](https://docs.meshai.dev/governance/quarantine.md): Quarantine unknown or suspicious agents and control their access through the proxy - [Agent Lifecycle Management](https://docs.meshai.dev/governance/agent-lifecycle.md): Set expiry dates, review schedules, and assign sponsors to enforce agent lifecycle governance - [Security Posture Score](https://docs.meshai.dev/governance/security-posture.md): A 0-100 score measuring your organization's AI agent security posture across 6 dimensions - [Attribute-Based Access Control (ABAC)](https://docs.meshai.dev/governance/abac.md): Assign owners to agents with granular permissions - control who can invoke, configure, and delete each agent - [Agent Relationship Graph](https://docs.meshai.dev/governance/relationships.md): Visualize agent dependencies and agent-to-agent delegation, auto-populated from proxy telemetry and OpenTelemetry span parentage - [HITL Approvals](https://docs.meshai.dev/governance/approvals.md): Human-in-the-loop approval workflow for AI agent requests - [Oversight Metrics](https://docs.meshai.dev/governance/oversight-metrics.md): Human oversight as measurable numbers: intervention demand, override rate, unattended approvals, and approval latency per agent - [Risk Classification](https://docs.meshai.dev/governance/risk-classification.md): EU AI Act risk levels for AI agents - classify, track, and govern - [Audit Trail](https://docs.meshai.dev/governance/audit-trail.md): Immutable audit log of every governance action - required by EU AI Act Article 12 - [Decision Evidence](https://docs.meshai.dev/governance/decision-evidence.md): The per-decision record MeshAI keeps when a policy evaluates an agent action - [EU AI Act Coverage](https://docs.meshai.dev/governance/eu-ai-act.md): Article-by-article mapping of EU AI Act requirements to MeshAI features - [Article 50 Transparency Evidence](https://docs.meshai.dev/governance/article-50-transparency.md): Record AI disclosure and content-marking evidence for EU AI Act Article 50 - [Fundamental Rights Impact Assessment](https://docs.meshai.dev/governance/fria.md): Article 27 FRIA - structured assessment of AI impact on fundamental rights - [Incident Reporting](https://docs.meshai.dev/governance/incidents.md): EU AI Act Article 73 serious incident reporting with deadline tracking - [Readiness Score](https://docs.meshai.dev/governance/readiness-score.md): 7-component compliance score (0–120) measuring EU AI Act readiness - [Plans & Pricing](https://docs.meshai.dev/billing/plans.md): Five tiers to match your AI agent deployment - from a free trial to enterprise - [Usage Limits](https://docs.meshai.dev/billing/usage-limits.md): Agent limits per plan and what happens when you reach them - [API Overview](https://docs.meshai.dev/api-reference/overview.md): Base URL, authentication, response format, and pagination for the MeshAI API - [Authentication](https://docs.meshai.dev/api-reference/authentication.md): API keys, scopes, signup, and login for the MeshAI API - [Agents API](https://docs.meshai.dev/api-reference/agents.md): Register, list, update, and delete agents. Manage teams. - [Telemetry API](https://docs.meshai.dev/api-reference/telemetry.md): Send heartbeats and report token usage - [Monitoring API](https://docs.meshai.dev/api-reference/monitoring.md): Fleet-wide status overview and per-agent health scores - [Cost API](https://docs.meshai.dev/api-reference/cost.md): Token spend attribution by agent, team, model, and time series - [Proxy API](https://docs.meshai.dev/api-reference/proxy.md): Proxy keys, managed provider keys, and proxy usage stats - [Anomalies API](https://docs.meshai.dev/api-reference/anomalies.md): List, inspect, acknowledge, and resolve anomalies. Manage anomaly rules. - [Governance API](https://docs.meshai.dev/api-reference/governance.md): Audit trail, risk classification, policies, and approval endpoints - [Security API](https://docs.meshai.dev/api-reference/security.md): Kill switch, quarantine, lifecycle, ownership, relationships, and security posture - [Compliance API](https://docs.meshai.dev/api-reference/compliance.md): Article 50 evidence, readiness, GPAI model inventory, FRIA, transparency cards, and incident endpoints - [Payload Capture API](https://docs.meshai.dev/api-reference/payload-capture.md): Configure opt-in, sanitized input payload capture - [Billing API](https://docs.meshai.dev/api-reference/billing.md): Query billing info, manage checkout, and handle webhooks - [Notifications API](https://docs.meshai.dev/api-reference/notifications.md): Manage notification channels for alerts and policy events - [Admin API](https://docs.meshai.dev/api-reference/admin.md): Platform-level tenant management (platform:admin scope required) - [Installation](https://docs.meshai.dev/sdk-reference/installation.md): Install the MeshAI Python SDK and configure optional dependencies - [MeshAI Client](https://docs.meshai.dev/sdk-reference/client.md): MeshAI class constructor and configuration reference - [Agents](https://docs.meshai.dev/sdk-reference/agents.md): Register, list, update, and delete agents via the Python SDK - [Telemetry](https://docs.meshai.dev/sdk-reference/telemetry.md): Heartbeats, usage tracking, and auto-tracking wrappers - [Cost Intelligence](https://docs.meshai.dev/sdk-reference/cost.md): Query token spend by agent, team, and model - [Anomalies](https://docs.meshai.dev/sdk-reference/anomalies.md): List, inspect, acknowledge, and resolve anomalies detected by MeshAI - [Governance](https://docs.meshai.dev/sdk-reference/governance.md): Risk classification, policies, approvals, and audit trail via the SDK - [Security](https://docs.meshai.dev/sdk-reference/security.md): Kill switch, quarantine, agent lifecycle, ownership, relationships, and security posture via the SDK - [Compliance](https://docs.meshai.dev/sdk-reference/compliance.md): Readiness score, FRIA, transparency cards, and incident reporting - [Billing](https://docs.meshai.dev/sdk-reference/billing.md): Query billing information and usage via the SDK ## OpenAPI Specs - [openapi](https://docs.meshai.dev/api-reference/openapi.json) This documentation is built and hosted on [Mintlify](https://mintlify.com), a developer documentation platform.