> ## Documentation Index > Fetch the complete documentation index at: https://docs.getmuster.io/llms.txt > Use this file to discover all available pages before exploring further. # What is muster? > AI agent observability and performance management for enterprise teams. muster is Elitery's managed AI operations service. Elitery deploys and operates the platform on your behalf — running discovery scanners, monitoring agent accuracy, tracking costs, and alerting your team when something needs attention. You connect your agents. Elitery does the rest. Get your first agent monitored in under 5 minutes. 3 env vars — works with any OTel-compatible agent. Connect n8n, Flowise, Bedrock, LangChain, and more. Precise business-rule checks via HTTP POST. ## Why muster? Most AI deployments in enterprises start the same way: one agent in Finance, another in Legal, a few in Operations. Within months you have dozens — many undiscovered, some with no owner, others running on outdated models. Elitery solves this with muster: * **Discover** — Elitery runs scanners and platform connectors across your environment, surfacing agents your team didn't declare * **Assess** — every agent is scored for risk automatically — PII exposure, external LLM calls, write access to systems * **Monitor** — Elitery tracks quality signal trends and alerts you when an agent's pass rate degrades * **Track** — per-agent costs calculated automatically from token counts and connector data, with department budgets and anomaly detection * **Report** — weekly health summaries from Elitery keep your team and leadership informed ## How agent data flows into Muster **OTel and platform connectors** cover discovery, cost tracking, and inferred quality signals — Muster's check inference engine automatically determines relevant checks from each agent's name and output structure. Zero code required. **The SDK adds precise checks** on top — for agents where inferred checks aren't enough. A bank's invoice processor needs `subtotal + tax == total` with the actual values, not a pattern match on the field name. The SDK lets your developers write those exact rules. Use it for Financial, Compliance, Legal, and Risk agents. Most customers start with OTel + connectors for full coverage, then add the SDK to their most critical agents. See [Integration guide →](/connectors/overview) ## How it works ``` Your Agents Integration muster Dashboard ────────── ─────────── ──────────────── invoice-agent (Python) → OTel env vars → Discovery · Quality · Cost fraud-agent (LangChain)→ OTel env vars → Anomalies · Heatmap n8n-workflow → Platform connector → Risk · Inventory custom-agent → HTTP quality signal → Recommendations ``` **The quality checks are worth writing regardless of muster.** Any agent in production should validate its own outputs. Muster turns those results into fleet-wide visibility — and Elitery monitors trends, alerts on degradation, and helps your team troubleshoot. See [What Elitery does for you →](/concepts/managed-service) for the full breakdown of responsibilities. ## Dashboard muster is accessible at [app.getmuster.io](https://app.getmuster.io). The dashboard includes: | Page | What it shows | How to get there | | ------------------ | ----------------------------------------------- | ---------------- | | **Executive** | Risk summary, spend, anomalies, signal coverage | Home | | **Inventory** | All agents — add agents manually here | Left nav | | **Discovery** | Scanner-found agents awaiting review | Left nav | | **Cost Dashboard** | Spend by agent, department, model | Left nav | | **Health Heatmap** | Quality signal pass rates across all agents | Left nav | | **Settings** | Benchmark sharing, instance config | Left nav |