# Keeping secrets from your AI agent

DevFeed: [Keeping secrets from your AI agent](<https://devfeed.tech/articles/keeping-secrets-from-your-ai-agent-17815.md>)

Original publisher: [Read original article](<https://encore.dev/blog/keeping-secrets-from-ai>)

Author: Ivan Cernja

Published: 2026-02-17T00:00:00Z

Content type: opinion

Language: en

Sources: [Encore Updates](<https://devfeed.tech/sources/encore-updates.md>)

Topics: [Security](<https://devfeed.tech/topics/security.md>), [Infrastructure as code](<https://devfeed.tech/topics/infrastructure-as-code.md>), [Architecture & Design](<https://devfeed.tech/topics/architecture-design.md>), [AI Agent](<https://devfeed.tech/topics/ai-agent.md>), [Back end](<https://devfeed.tech/topics/backend.md>)

Tags: [agent](<https://devfeed.tech/tags/agent.md>), [ai](<https://devfeed.tech/tags/ai.md>), [architecture](<https://devfeed.tech/tags/architecture.md>), [backend](<https://devfeed.tech/tags/backend.md>), [ci-cd](<https://devfeed.tech/tags/ci-cd.md>), [docker](<https://devfeed.tech/tags/docker.md>), [infrastructure-as-code](<https://devfeed.tech/tags/infrastructure-as-code.md>), [llm](<https://devfeed.tech/tags/llm.md>), [security](<https://devfeed.tech/tags/security.md>), [terraform](<https://devfeed.tech/tags/terraform.md>)

## AI overview

The article argues that AI coding tools can expose committed infrastructure configuration and sensitive local files, while file-exclusion mechanisms are not reliable security boundaries. It presents infrastructure-from-code as an approach that keeps credentials and connection details out of repositories by resolving them at runtime.

## Source excerpt

Your backend architecture determines what AI tools see, send, and write.