Not on purpose, but because you don’t actually know what they’re running right now. Scripts change. Settings drift. Deploys get hotfixed in ways no one documents. The “configuration” you think you have in production? It’s probably different from what’s in your repo. This is how small bugs turn into outages, how compliance audits fail, and how security holes open wide without anyone noticing.
Agent Configuration Policy-As-Code is the antidote. It turns ephemeral, invisible agent settings into something concrete, version-controlled, and automated. No more manually syncing YAMLs, SSH’ing into boxes, or guessing whether your agents match your intended state.
When configuration becomes code, everything changes. You can:
- Define all agent policies in a single source of truth.
- Track every change and rollback instantly when something breaks.
- Enforce configuration baselines across all environments with zero manual intervention.
- Apply the same review and approval process you already use for application code.
This approach hits every pain point at once. Drift disappears. Compliance stops being a scramble. Security hardens by default. Onboarding new environments becomes a git clone away from done.