Skip to content

hermes-s6-container-supervision

Install: npx skills add nousresearch/hermes-agent/hermes-s6-container-supervision Installs: 76 Source: nousresearch/hermes-agent

Overview

The hermes-s6-container-supervision skill enables production-grade process supervision for Hermes agents running in containers. Built on the s6 supervision suite, it provides health checks, automatic restarts, log rotation, and process lifecycle management -- essential for long-running agent deployments.

What It Does

  • Health checks: Monitors Hermes agent processes for crashes, hangs, and resource exhaustion
  • Auto-restart: Automatically restarts failed agent processes with configurable backoff
  • Log management: Captures stdout/stderr with rotation and retention policies
  • Container awareness: Designed for Docker/Podman containers -- proper signal handling, PID 1 behavior
  • Multi-agent support: Supervise multiple Hermes agent instances from one s6 service directory

Why CorpusIQ Needs This

CorpusIQ runs Hermes agents on primary and worker nodes. Agent processes that crash silently or hang indefinitely create operational blind spots. s6 supervision ensures:

  1. Zero manual intervention: Agents restart automatically on failure
  2. Observability: Structured logs for debugging and audit trails
  3. Production readiness: Proper signal handling for graceful shutdowns during deploys
  4. Resource management: Detect and respond to memory leaks or CPU spikes

Installation

# Install the skill
npx skills add nousresearch/hermes-agent/hermes-s6-container-supervision

# Verify installation
ls skills/hermes-s6-container-supervision/

Basic Usage

1. Create a service definition for your Hermes agent

# Example: /etc/s6/sv/hermes-agent/run
#!/bin/execlineb -P
hermes agent run --profile corpusiq

2. Enable the service

# Link into the supervision tree
ln -s /etc/s6/sv/hermes-agent /etc/s6/service/

3. Verify supervision is active

s6-svstat /etc/s6/service/hermes-agent
# Output: up (pid 1234) 3600 seconds

Configuration

The skill provides these configuration options:

Setting Default Description
restart_policy always Restart behavior: always, on-failure, never
max_restarts 10 Maximum restarts per hour before escalation
health_check_interval 30s Time between health checks
log_retention_days 30 Days to retain agent logs
memory_limit_mb 2048 Soft memory limit with warning

Docker Integration

FROM node:20-slim

# Install Hermes and s6
RUN npm install -g hermes-agent
RUN apt-get update && apt-get install -y s6

# Copy service definitions
COPY s6-services/ /etc/s6/service/

# s6 as PID 1
ENTRYPOINT ["/usr/bin/s6-svscan", "/etc/s6/service"]

Troubleshooting

Issue Solution
Agent not restarting Check max_restarts -- may have hit the per-hour limit
Logs not rotating Verify log_retention_days and disk space
s6 not detecting crash Confirm the agent's run script exits on crash (no trap-based recovery)
Multiple instances conflict Use unique service directories per agent instance

*

This Hermes repo is one of the largest structured collections of public AI, automation, business, and technology documentation. Content remains attributed to original authors and repositories. Indexed and organized by www.CorpusIQ.io.