Skip to content

getting-started.md

Latest commit

 

History

History
122 lines (88 loc) · 3.22 KB

getting-started.md

File metadata and controls

122 lines (88 loc) · 3.22 KB

Getting Started

Install and configure TelemetryFlow Hermes for autonomous incident response.

Prerequisites

Requirement Version Check
Python 3 3.8+ python3 --version
Hermes Agent Latest hermes --version
TelemetryFlow Platform Running curl $TELEMETRYFLOW_API_URL/health
ClickHouse Running Via TFO API
Telegram Bot Token 4 tokens @BotFather
kubectl 1.28+ kubectl version

Installation

Step 1 — Clone TelemetryFlow Hermes

git clone https://github.com/telemetryflow/telemetryflow-hermes.git
cd telemetryflow-hermes

Step 2 — Setup Environment

make env

Edit ~/.hermes/.env with your credentials. See Environment Variables for the complete reference.

Minimum required variables:

# TFO Platform Authentication (pick one method)
TELEMETRYFLOW_API_KEY=tfs_your_api_key_here

# TelemetryFlow Platform Connection
TELEMETRYFLOW_API_URL=http://localhost:3000/api/v2
TELEMETRYFLOW_ORGANIZATION_ID=your-org-uuid
TELEMETRYFLOW_WORKSPACE_ID=your-workspace-uuid

# LLM Provider (for Hermes agent reasoning)
ZHIPU_API_KEY=your_zhipu_key         # Triage/Reviewer/Remediator
ANTHROPIC_API_KEY=your_anthropic_key  # Investigator

Step 3 — Create Telegram Bots

Create 4 bots via @BotFather on Telegram. Each agent needs its own bot token (Telegram allows 1 connection per token).

Step 4 — First-Time Setup

make init

This installs the Hermes agent and configures everything:

  • 4 agent profiles (triage, investigator, reviewer, remediator)
  • 29 skills across 18 categories
  • 6 cron jobs
  • 40 plugin tools covering all 20 TFO Platform modules
  • 3 lifecycle hooks
  • ClickHouse read-only security
  • 4 Telegram gateways

Step 5 — Verify Pipeline

make verify

This runs end-to-end verification:

  • Hermes installation
  • API connectivity
  • ClickHouse access
  • Telegram gateway
  • Tool availability

Step 6 — Start Gateways

make deploy

Starts all 4 Telegram gateways in background.

First Investigation

Send a test alert to your Triage agent's Telegram bot:

ALERT: payments-api p95 latency breach
Service: payments-api
Metric: http_server_duration_p95
Value: 640ms (threshold: 200ms)
Severity: HIGH
Time: 2026-06-04T03:47:00Z

The agent pipeline will:

graph LR
    A["Alert received"] --> B["Triage classifies"]
    B --> C["Investigator gathers evidence"]
    C --> D["Reviewer verifies"]
    D --> E["Remediator proposes action"]
    E --> F["Human approves"]
Loading

Expected timeline: ~23 seconds from alert to proposed remediation.

Next Steps