config.yaml

# Available models from Ollama:
# - mistral
# - llama2
# - codellama
# - phi
# - gemma
# - tinyllama
# - dolphin-mixtral
# - or your own custom pulled model name
model: mistral
source: journalctl  # or "file", "docker"
log_file_path: /var/log/syslog  # only used if source == file
priority: err       # journalctl log level: emerg, alert, crit, err, warning, notice, info, debug
entries: 500       # number of log entries to fetch
timeout: 90         # number of seconds before LLM timeout
docker_container: my_container # this can be a name or an id like 'f77ec9ab2112'
ollama_host: http://localhost:11434 # if you're running a remote ollama, change this.
prompt: |
  You are a helpful Linux operations assistant. Analyze the following logs:
  - Identify root causes
  - Summarize key issues
  - Recommend next steps

  LOGS:
  {{LOGS}}