SessionSleuth: forensic audit tool for AI coding agents' sessions
SessionSleuth tackles the lack of incident‑ready logs for AI coding assistants by turning Claude Code and Codex artifacts into a deterministic, verifiable timeline. It offers forensic and gate modes that run locally without external services, plus delegation review and session audit commands that use existing agent CLIs. The tool is aimed at developers and security teams who need to ensure AI‑generated code is trustworthy and auditable. Compared to ad‑hoc scripts, it provides structured evidence, security flags, and reproducible reports out‑of‑the‑box.
View on GitHub →Marker-Inc-Korea/sessionsleuth