Skip to main content

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.

​
1. Sign up

Go to app.getmuster.io and create your account. Your 14-day free trial starts immediately — no credit card required.

​
2. Add your agent

From the dashboard, click Add Agent and fill in:
  • Name — e.g. invoice-processor-v2
  • Framework — LangChain, LangGraph, n8n, etc.
  • Owner email and Department
muster assigns a risk score automatically based on what your agent does.

​
3. Connect your agent

Choose the path that fits your agent: Option A — OTel (recommended for coded agents): Set three env vars and Muster receives traces, infers quality checks, and tracks costs automatically. 
OTEL_EXPORTER_OTLP_ENDPOINT=https://backend.getmuster.io/api/v1/otel
OTEL_EXPORTER_OTLP_HEADERS=Authorization=Bearer <your-jwt>
OTEL_EXPORTER_OTLP_PROTOCOL=http/json
Option B — Platform connector (for n8n, Flowise, Dify): Elitery deploys a connector that monitors your flows automatically. No action needed from you. Option C — SDK / precise checks (for Finance, Compliance, Legal agents): Install muster-sdk to run exact business-rule checks against each output. 
from muster_sdk import MusterAgent
agent = MusterAgent(agent_id="invoice-processor-v2", muster_url="...", token="...")

@agent.govern(required_fields=["vendor_name", "total"])
async def process_invoice(input_data): ...
Or send a plain HTTP signal if you prefer no library:

​
4. Send a quality signal (Option C / plain HTTP)

Send signals to your backend URL, not the dashboard URL.
  • Trial / shared instance: https://backend.getmuster.io
  • Your dedicated instance: You’ll receive your own URL from Elitery (e.g. https://backend.yourcompany.getmuster.io)
The URL is also shown in your dashboard → Onboarding page.
If using Option C (plain HTTP), add this after your agent runs. No library required. 
import httpx, threading, time

def muster_emit(job_id, checks, token_input=None, token_output=None, model=None, latency_ms=None):
    """Fire-and-forget. One call covers correctness + cost + latency."""
    def _send():
        try:
            httpx.post(
                f"https://backend.getmuster.io/api/v1/jobs/{job_id}/quality",
                json={
                    "agent_id": "invoice-processor-v2",
                    "job_id": job_id,
                    "overall_passed": all(c["passed"] for c in checks),
                    "checks": checks,
                    "token_input": token_input,   # → cost tracking
                    "token_output": token_output,
                    "model": model,
                    "latency_ms": latency_ms,     # → SLA monitoring
                },
                timeout=2.0,
            )
        except Exception:
            pass  # never block your agent
    threading.Thread(target=_send, daemon=True).start()

# Call after your agent finishes each job
start = time.time()
output = your_agent.run(input)  # your existing code

muster_emit(
    job_id=job_id,
    checks=[
        {"check_id": "output_not_empty",    "severity": "HIGH", "passed": bool(output)},
        {"check_id": "subtotal_arithmetic", "severity": "HIGH",
         "passed": abs(computed - declared) < 0.01,
         "expected": str(declared), "actual": str(computed)},
    ],
    token_input=output.usage.prompt_tokens,      # from your LLM response
    token_output=output.usage.completion_tokens,
    model="gpt-4o",
    latency_ms=int((time.time() - start) * 1000),
)

​
5. View your data

Head to the Health Heatmap in your dashboard. Within minutes of your first signal you’ll see:
  • Pass rate per check (e.g. subtotal_arithmetic: 94%)
  • Trend direction — Stable / Degrading / Improving
  • Cost data if you included token_input / token_output
The quality endpoint requires no authentication. Rate limit: 1,000 requests/minute per agent.

​
What to check

Pick checks that reflect your agent’s actual job. A few starting points:
Agent typeGood first checks
Data extractionoutput_not_empty, required_fields_present, subtotal_arithmetic
Classificationdecision_is_valid_enum, decision_has_rationale
Document reviewclause_identification, risk_flag_present
Any agentoutput_not_empty, latency_within_sla
See the full check ID reference for more.

​
Next steps

OTel integration

3 env vars — works with LangChain, CrewAI, or any custom agent.

Platform connectors

Elitery-managed connectors for n8n, Flowise, Bedrock, and more.

SDK guide

Precise business-rule checks for Finance, Compliance, and Legal agents.

Quality checks concept

How muster tracks trends and detects degradation.