Painted Wolf Code is an Apache-2.0, local-first AI coding agent built in Go and specifically tuned for open source models. Designed to provide greater transparency than SaaS alternatives, it operates as a workflow engine using declarative YAML and composable templates for agent personas and tool guidance. The system emphasizes reliability and security through a durable evidence ledger that grounds agent output with verified citations, and integrated security scanning—including SCA, secrets detection, and SAST—which is wired directly into workflow gates. This session introduces a local-first approach to AI coding where orchestration is tuned for open source models, prioritizing deep customization and evidence-based verification.
Target Audience: Software developers interested in AI-assisted coding