AI agent security
Prompt injection, skill supply chains, secrets and the gates around agents.
Posts tagged AI agent security
AI agent skill security: three CI gates before a skill ships
AI agent skills are a software supply chain. Three CI gates (an LLM scanner, static analysis, a duplicate check) and how to keep them failing closed.
GitHub OIDC and Terraform: short-lived model credentials in CI
Give a CI job model access without storing a key: GitHub Actions OIDC, an AWS IAM role in Terraform, a tight sub condition, and the pitfalls to avoid.